IMAGING SERVICES NORTH Boston Spa, Wetherby West Yorkshire, LS23 7BQ

www.bl.uk

This PDF was created from the British Library’s microfilm copy of the original thesis. As such the images are greyscale and no colour was captured.

Due to the scanning process, an area greater than the page area is recorded and extraneous details can be captured.

This is the best available copy D67043ÌB6 \ l ì \

Attention is drawn to ^ the fact that the È copyright o f this thesis rests with its author. This copy o f the thesis has been supplied on condition that anyone who consults it is p- understood to recognise that its copyright rests with its author and that no quotation from the thesis''and no information derived from it may be published without the author’s .pri^^ written consent.

- if

r>r»FN,..ii Iimm f -pm '

r/. ' Hi.

REPRODUCED FROM THE

AVAILABLE COPY NEW ASPECTS OF THE MICHAELIS-ARBUZOV AND PERKOW REACTIONS

A Thesis SubniittGd. Top the Degpee of*

DOCTOR OP PHILOSOPHY

of the

COUNCIL FOR NATIONAL ACADEMIC AWARDS

by

Mrs Lubomyra POWROZNYK MSc (Lublin)

THE POLYTECHNIC OP NORTH LONDON Holloway Road, London, N? 8 DB October 1985 ACKNOWLEDGEMENTS

The authoress owes a great debt and wishes to

express her deep gratitude to her supervisor.

Dr, H, R. HUDSON, for liis constant guidance, kind patience and helpful discussions during

this long investigation.

Thanks are also extended to Dr. R. w. Matthews and Mr, J. Crowder for assistance with F and

C n.m.r. measurements, and Dr. K. Henrick for

X>ray diffraction studies. DECLARATION

While registered as a candidate for the degree of Ph.D.

for which this subraission is made, I have not been

registered as a candidate for any other award.

In partial fulfilment of the requirements for the

degree, an evening course of MSc lectures on Structural

Methods (uv, ir, nmr, mass spectometry and X-ray crystallography) and research colloquia have been attended in the School of Chemistry,

Mrs L. Powroznyk

ABSTRACT

NEW ASPECTS OF THE MICHAELIS-ARBUZOV AND PERKOW REACTIONS

Luborayra Powroznyk

The work presented in this thesis has involved an inves- gatxonof the thermal decomposition of the Michaelis (o (PhO)5 PMeBr, (PhOjPMel,

been shojm to occur by nucleophilic substitution in L e

In addition, rnovel^’difpioportionation reactiorgave^tht^™’ 2n r t S quasiphosphonium intermediates and the tetrampfhvi rbrx vx n i ^ salt Which were obtained rrom dispropor by ■>H''"™ p ^ a n d ° §r""® Isolated and identified y n, 1 P, and 1 n.m.r. spectroscopy. In certain C0n t ^ n i n r r p % ® p ‘'?-'‘;i“ ‘*^ proportion of an intermediate -r fu ^ ® linkage was also observed. The mechanisms of these reactions and the spectroscopic features of t h ^ tud T h ^ r S ‘'p n L r ‘^d°"/‘’^ en/produdrhLfbe^e: ^ P-O-P 1 intermediate was isolated in very small amount and was found to be stable in deuterochloroform in a sealed tube, but not as a solid. «^^erocnioroform in

hindered phosphorus (III) esters viy (R0 )3 P, (R0 )2 PPh, and R0PPh2 (R = Me3 CCHo), th^f^rst examples of identifiable intermediates ?ROU p+ph p h L r Ph2 Pi(0R)CH2 cSphB;-, ani "^COPhBr-, Ph2 PMORloct:CH2 )PhCl- L v e been^obtained ?n the reactions bv intermediates have been investig^Ud ^ H n.m.r. spectroscopy and the mechanisms^nC their decomposition to Arbuzov and Perkow uroducts hot/o been elucidated. The intermediates PhP+(OR)oGHorr>Pv r - rnriCDir^°L^'^? 2CpPhBr: were suaoiestable at a? roomr o o m temp ' S e r ^ t u r " f L CDCICDOI5.. heating in sealed tubes (O-R(0.5 h )i atq ^-^i o O ^ c ) 3 the dilute solutions decomposed completely to sive nnn^ products of PhP(0)(0R)CH2 C0Ph and of PhoPf0 )cLrnPh respectively. Neopentyl gromide was f S i ^ L c h case. studies on the réaction*^ nc \ 1-

have also revealed a competing process of P-N cleavaee o?\hfh"îid: p“î:::^nt!" the'=^:?d^nf:s

1.1 MICHAELIS-ARBUZOV REACTION

Tlie interaction of trialkyl phosphites and alkyl hali-

des, referred to as the Michaelis-Arbuzov reaction, was

first reported by Michaelis^ and later exploited extensi-

vely by Arbuzov'^ and many other workers.

(RO)^P + R'X (R0)2P(0)R' + RX

A stable 1:1 intermediate was obtainable in certain

instances from triaryl phosphites^-® and in other special

cases involving trialkyl phosphites and a,p-dihaloalkyl ethers, ^ ^

The first step in this type of reaction was made by

Michaelis and Kahne, who reported in 1898 that the reac­

tion of triphenyl phosphite with iodomethane at 100 °C for

48 h yielded 1 :1 adduct as crystalline needles, m.p. 70- 75

(C^H^O)^P + CH^I (CgH^O)^PCH^I

Adducts of this type are generally known as quasiphos- phonium intermediates * ^ and are readily hydrolysed to give the phosphonate, phenol, and hydrogen iodide.

(C6H^0)^PCH^I + H*0H — ^ HI + C^H^OH + CH^P0(0C^H^

Michaelis and Kahne prepared a number of other interme­ diates e.g. from tri-m-tolyl or tri-p-tolyl phosphite and iodomethane by heating the reactants together on a water bath, Tri-ni-tolyl phoaphite waü thus heated for 5 h to

give a dark brown oily liquid, which became crystalline

after cooling, and after washing with dry ether it gave a

hygroscopic crystalline product.^

The product from tri-p-tolyl phosphite, obtained after

12 h, remained however as a viscous liquid.^*

Triphenyl phosphite and benzyl chloride gave a syrupy

liquid on heating at 150-175 °C, instead of the expected

crystalline phosphonium compound,

A number of aromatic intermediates were also produced

by Arbuzov in 1906, but both Michaelis and Arbuzov failed

to obtain intermediates from trialkyl phosphites and alkyl

halides. They showed that the final products in these

reactions were usually phosphonates and alkyl halides,

although Michaelis and Kahne obtained methylphosphonic

acid, ethyl iodide and ethene by heating triethyl phosphi­

te with methyl iodide at 110-220 °C for 12 h. (G2H^0 )^PCH^I > J CH^POiOH)^ + C^H^I +

They assumed from the beginning that a phosphoniuia

intermediate was involved:

(RO)^P + R»X -* 1(r o ),p r ':^

The thermal decomposition of triphenyl phosphite-alkyl

halide adducts was investigated by Arbuzov^^ in 1905, by

heating the compounds in sealed tubes at 210-220 °C. Some

tar formation occurred, especially with the higher mem­

bers, along with free iodine, and the following decomposi-

tion products were isolated. Temp. Time Yield (%) Adduct n (h) Phi RP(0)(0Ph)2

(PhO)^PMel 220 20 74 56 ^ (PhO)^PEtl 220 20 71 60 - (PhO)-PCH. ,PhI 208 20 J 2 15

a 196-9 °C, 1.5520

b b ^2 200-4 °C, 1.5454

The isobutyl iodide adduct, however, gave only iodine

as the isolated product after 15 h at 210 °G.

The Michaelis-Arbuzov reaction^^ takes place with any

derivative of trivalent phosphorus that carries an ester group (OR):

> - 0 R + R»X — > 1P(0R)R» X — ► P(0)R' + RX

and most varied selection of organic halides R ’X are able

to participate in the reaction. The chief requirement of

the structure of the organic halide is that the halogen

atom must be capable of nucleophilic displacement, this occurring most easily if it is attached to the terminal carbon atom of an aliphatic chain. It must not be directly connected to an aromatic ring.

The above reaction is merely one instance of the more general Michaelis-Arbuzov rearrangement, ^ or isomerisa­ tion. If the radical R» of the alkyl halide is identical with the radical R of the ester group the reaction takes on the aspect of a true isomerisation,^^ which can be performed with minute amounts of the halide R'X. Landauer and Rydon thus shoived that isomerisation^ of trimethyl phosphite occurred with methyl iodide (0.1 mol)

by heating the reactants under reflux at 100 °C. After a

brief induction period, a vigorous reaction set in and

was complete within 5 min. Distillation then gave methyl iodide and dimethyl methylphosphonate:

(MeO)^P + Mel (Me0)2P(0)Me + Mel

Although customarily used with tertiary phosphites,

(RO)^P, from which the reaction product is an ester of a

primary phosphonic acid R»P(0)(0R )2 i.e. a PHOSPHONATE,^^

the reaction may also be carried out with esters of phos-

phonous acids, RP(0R)2, in which case esters of secon- ^ r y phosphonic acids R»RP(0)0R, i.e. PHOSPHINATES are formed.

RP(0R)2 + R»X — ^ R'RP(0)0R + RX

Esters of phosphinous acids, R^POR, form tertiary phosphine oxides R^P=0 as follows:

R^POR + R»X R2R’P=0 + RX

The reactivity of phosphorus (III) esters increases from phosphites to phosphonites and again to phosphinites, in accord with a general increase in reactivity upon approach to the tertiary phosphine structure.^

Although the reaction is frequently conducted by mixing the reactants and warming the mixture to the required tem­ perature,'*^ most often IOO-I5O °C, the most reactive combi­ nations enter the reaction at room temperature, this being especially true of reactions with acyl halides. Crystalline quasipliosphonium salts^^ have in some

cases been prepared by the addition of methyl iodide to

unsaturated phosphonous diesters containing branched

ester radicals, such as GH2=CH-CH2 P(OBu^)^, although the

adducts decompose to secondary phosphoric esters on mild heating; 0 CH2=CH-CH2P'"(0Bu ^)^RX~ — » CHp= CH-OBu^ + Bu^X

The phosphonous diesters are also more susceptible to

self-isomerisation than the trialkyl phosphites, but in a

truly pure state they isomerize with difficulty. Esters

of aromatic phosphonous acids isomerize at 250 ^C, but the aliphatic compounds are stable to 300

Aromatic esters^° of phosphonous acids interact with alkyl halides to form extremely stable quasiphosphonium salts that are not decomposed under the normal conditions of the Michaelis—Arbuzov reaction.

The salts can, however, be decomposed by drasting hea- ting or by alcoholysis^^ or hydrolysis. The thermal decomposition has further been shown to be complicated by redistribution of the alkyl and phenoxy groups in the pro­ ducts that are obtained.

8 1.2 QUASIPHOSPHOlliUM COMPOUNDS IN TflE PREPARATION OF ALKYL HALIDES

Landauer and Rydon in 1953 reported a new method for

the preparation of alkyl halides from phosphonium inter­

mediates and alcohols.® They heated triphenyl phosphite

with methyl iodide under reflux to obtain the methiodide,

(PhO)^PMel, and in an attempt to determine the nature of

the phosphorus-iodide bond in this compound, they treated

it with alcoholic silver nitrate. Precipitation of silver

iodide was slow and not instantaneous as with methyltriphe-

nylphosphonium iodide.

In view of the relatively slow rate at which silver

iodide was formed from the triphenyl phosphite methiodide

the possibility of a penta-coordinate structure for the

phosphorus compound was accepted.

In order to exclude the possibility that the time - de­

pendent production of silver iodide was due to reaction of

the silver nitrate with methyl iodide formed by dissocia­

tion, the compound was treated with cold ethanol and methyl iodide was sought in product. Surprisingly enough, ethyl iodide was formed together with phenol and diphenyl methylphosphonate, thus:

(PhO)^PMel + ROH — > RI + PhOH + (Ph0)2P(0)Me

The experiment thus led to a new method for the prepara­ tion of alkyl iodides.

The reaction appeared to be quite general and gave excGllent yields of sikyi iodides froiu a wid© variety of

saturated and unsaturated alcohols, including cholesterol,

p;lycols and hydroxyesters. Landauer and Rydon also found

that it was not necessary to isolate the methiodide and

that alkyl chlorides could be prepared similarly from

triphenyl phosphite and benzyl chloride.

Hydrogen halides were also shown to react analogously.

RÜH + (PhO)^P + HHal RHal + PhOH + (PhO)2 p(0 )H

Alkyl halides may thus be prepared by a general proce­

dure which involves refluxing a mixture of an alcohol,

triphenyl phosphite, and an alkyl, hydrogen, or other

halide, thus: ROH + (PhO)^P + R'Hal — RHal + PhOH +

(PhO)2 P(0 )R» (R» = alkyl, H, NH^, Li, Na).

The new method gave excellent results with sterically

hindered alcohols which are difficult to convert into

halides by conventional methods, i.e. neopentyl iodide was

obtained from the alcohol in 74% yield.

1.3 TRIARYL PHOSPHITE DIHALIDES AND THEIR DISPROPORTIONATION

In 1954 Landauer and Rydon^^ prepared triaryl phosphite dihalides, (ArO)^PHal2 , as crystalline solids. Triphenyl phosphite dichloride, dibromide, di-iodide, bromo-chloride bromo-iodide and chloro-iodide were obtained, the dichlo­ ride being obtained as analytically pure, colourless prisms, m.p. 80-81 °C. Earlier workers had described most of these substances as oils or unstable liquids.

10 The triphenyl phosphite dihalides were shown to react

very readily with alcohols to form alkyl halides, which

were assuined to be formed in an Arbuzov-type reaction by

halide ion attack on the ester - exchanged intermediate.

(PhO)^PX2 (PhO)2 P'^ PhOH

(PhO)2 P(0 )X RX

A simpler procedure was by the direct interaction of the

three components:

(PhO)^P + ROH + Hal^ RHal + PhOH+ (Ph0 )2 P(0 )Hal

However in attempts to prepare highly purified specimens

of the triphenoxyphosphorus dihalides for conductivity and

X-ray investigations, Rydon and Tonge found that these

compounds were less stable and more complex in structure

than had been supposed.^® For example, on repeated recrys­

tallisation, triphenyloxyphosphonium dibromide (PhO) PBr 3 2 yielded tetraphenoxyphosphonium bromide (PhO)^PBr.

Further investigation showed also that two molecules of

triphenyl phosphite reacted with one molecule of halogen

in chlorobenzene solution to produce diphenyl phosphorobro-

midite (PhO)2 PBr, and the tetraphenoxyphosphoniuia bromide

(PhO)^PBr, which could be isolated as a crystalline solid,

2(PhO)^P + Hal2 (PhO)2 PHal + (PhO)^PHal

Exactly similar results were obtained with chlorine, and iodine.

The mechanism of reaction seems to be^^ ^hat the dihalide is first formed and then exchanges one of its halogens for

11 a phenoxy group from the second phosphite molecule:

(PhO)^P + Hal, (PhO)xPHal^ j 2 (PhO)^PHal., + (PhO)^P (PhO)^PHal + (PhO)2 PHal

There is some evidence that the reaction between chlo­

rine and diphenyl phosphorochloridite involves an analo­ gous stage.

2(PhO)2PGl + Cl^ — > (PhO)^PCl2 + PhO-PGl^

In agreement with the findings of Hari’is and Payne,

there is no increase in the number of ions when further ha-

logen is added in a solvent of sufficient ionising power,

and the reaction may take the following course:

(PhO)^P'^ + Hal" + (PhO)2 PHal + Hal^ — ^

[(H.O),p] ,,.PHal, 2 4j

In a non-ionising solvent this second stage may involve covalent species:

(PhO)^PHal + (PhO)2PHal + Hal^ — ^ (PhO)^PHal + (PhO)2PHal^

Halogen transfer may then give rise to the ions

[(PhO)^pJ and [(PhO)2 PHal^]’ or the exchange of halogen and pherioxy may give the covalent form of the triphenoxy- dihalide (PhO^PHal^.

The mechanism previously advanced^^»® for the action of the phenoxy-halides on alcohols requires modification in the light of the dimeric ionic structures.

The reactions of compounds of the general type KJ" with alcohols are most reasonably regarded as involving

12 attack by the oxygen atom of the alcohol on the phospho­

nium cation, and attack on the alkyl group of the alcohol

by one of the nucleophilic halogen atoms of the anion.

These processes may be simultaneous or sequential:

0-R Y-P“Y^ . X^P-OH + RY + PY^ or H

X. P^ 0-R X,P-0‘*’-R 4 I H

X.P-O'^-R Y-P’Yj- - X^P-OH + RY + PY^ ^ I D H

The process will be completed by loss of HX from the

hydroxylated cation: X^P-OH X^PO + HX

Alcoholysis of the dimeric forms of the tetraphenoxy-

monochloride and monobromide, may occur in two steps as follows:

[(PhO)^p]'^ [(PhO)^PHal2] ” + ROH — ^

(PhO),PO + PhOH + RHal + (PhO).Hal

(PhO)^Hal + ROH RHal + (PhO)^PO + PhOH

The overall reaction is thus:

[(PhO)^p]'^ [(PhO)^PHal2 ]" + 2R0H — ».

2(PhO)^PO + 2 PhOH + 2RHal

The disproportionation reactions of the triaryl phos­ phite dihalides probably occur via a series of exchange reactions between lialide and aryloxide ions, involving phosphonium and phosphorane structures.

(ArO)^P'^Cl Cl” ^ (ArO)^PCl2 ^ (ArO)2 P‘^Cl2 OAr (ArO)^P+Cl OAr“ (ArO)^PCl ^ (ArO)^P”^ C1‘

13 A whole range of salt a of the general type RphO) PX "1 r , L n 4-n J i(PhO)^^PXg^J can tJius be formed.

1.4 disproportionation o f quasiphosphonium s a l t s

In 1967 Nesterov reported disproportionation of methyl

and phenoxyl groups in the decomposition of certain quasi-

phosphonituri salts.^^ He thus showed that the thermal de­

composition of MePh^P(OPh)^_^I (n = 1 or 2) was accompa­

nied by a restribution of Me and PhO groups, as a result

of which the products containing the phosphoryl group were

freed of methyl groups, the other reaction product being

the salt of a tetra-substituted phosphonium ion. Equimolar

mixtures of Mel and PhPCOPh)^ or of Mel and Ph^PiOPh) were

heated in sealed tubes to give the crystalline quasiphos­

phonium salts. Further heating in PhN02 vapour (208 °C)

then brought about a slow decomposition of the diphenoxy

derivative to give three main products: iodobenzene,

diphenyl phenylphosphonate, and phenyltrimethylphosphonium

iodide, together with unidentified materials (43%).

MePhP(0Ph)2 l — 2-^ phi + (Ph0)pP(0)Ph + PhPMe^I

56 55% 7 1 .5% hours 21%

Similar heating of MePh^PiOPh)! gave little decomposi­

tion even at 550 °C, hence the salt was heated in a flask with an open flame yielding a distillate of iodobenzene. phenyl diphenylphosphinate and diphenyldimethylphosphonium iodide :

14 MePh^PiOnOl > Phi + Ph P(0)0Pli + Me Hi PI

95%

The possible ifiechariisifis by which these processes

occurred were not, however, discussed.

An anoinalous reaction between triphenyl phosphite and

laethyl bromide, discovered by Weekes in 1972, led to for­

mation of the unexpected dimethyldiphenoxyphosphoniura

bromide (PhO)2 P'^Me2 Br~, and suggested that a similar

disproportionation might have occurred.

The product was obtained when triphenyl phosphite was

heated with an excess of methyl bromide at ca. 180 ^C.

Then, after removal of the volatile materials (mainly

bromobenzene and a small amount of methyl bromide) the

H n.m.r. spectrum of the residue showed the presence of

two phosphorus compounds by virtue of sharp signals for

two P-Me doublets in the aliphatic region at ¿” 2.69 and 1 .77.

Anhydrous ether was then added to the residue to give

a white solid which was recrystallised from Analar acetone

and identified as dimethyldiphenyloxyphosphonium bromide,

(PhO)2 P Me^Br , by elemental analysis and n.m.r. spectro­ scopy. 1 H n.m.r. spectrum in deuterochloroform showed! 7 .0

(Me doublet, 3H, 14.6 Hz);!'2.48 (phenoxy protons, singlet, 5H) and the ^'’p spectrum (CDCl^) showed a multi­ plet, S' 95 ppm.

The evidence suggested that at elevated temperatures and in the presence of an excess of methyl bromide an extra

1.5 methyl group can be combined into the phosphonium raolecule

at the expense of the loss of a phenoxy group. No further

investigations of these processes have, however been carried out,

1.5 PERKOW REACTIONS

Trialkyl phosphites are known to react with cr-halogeno-

carbonyl compounds to give two products the proportions of

which depend on the nature of the halogen and the reaction 37 conditions. The usually accepted mechanisms for these

reactions are as follows:

(1) Perkow Reaction

Ph (RO)^P: (ROjP'-Çç-O- CH^Cl CH^~ Cl.

Ph 0 Ph / , , II / (RO)^P'^— 0~ C C1‘ (R0)2P— 0-C + RCl 5 \ \ CH2 ^CH^ yinyloxyphosphonium vinyl phosphate intermediate

The essential characteristic of this reaction is the formation of a vinyloxyphosphonium intermediate which under- goes dealkylation to form the definitive vinyl phosphate.

(2) Arbuzov Reaction

(RO),P:^CHjr^Br — (RO),P'^-CH„ Br (RO)^P-CH^ -f. RBr 3 I 2 3 I 2 2 I 2 C=0 I Ph ketophosphonium ketophosphonate intermediate

1.6 This reaction is characterised by the formation of a

ketophosphonium intermediate which also undergoes dealky­

lation to form the phosplionate product.

The essential difference between the two reactions is

the manner in which the initial attack of the phosphorus

atom on the oc-halogenoketone takes place.

In the Perkow reaction electrons are donated by the

phosphorus atom to the carbon atom of the carbonyl group

to form a betaine which then undergoes rearrangement so

that a halide ion is released with the formation of a vinyl group.

In the Arbuzov reaction, the electrons are initially

donated to the ci-CH^ group and a halide ion is released directly.

In the both reactions a phosphoniuia intermediate is

formed and this subsequently undergoes dealkylation to

yield the corresponding products. Other modes of reaction

may also be possible in certain circumstances, i.e. attack

by phosphorus on the halogen atom to give a halophosphonium

enolate which then rearranges to give either the ketophos-

phonium or vinyloxyphosphonium intermediate. A sequence of

this type has been proposed in reactions involving ex, cx-di- bromoketones and in those cases in which the carbonyl car­ bon atom is difficult to attack because of steric hind­ rance .

Br 0 . I (RO)^P BrCHCOAr (RO)^P'^Br

17 Ar (RO)^P +— 0 or (RO)^p-*-— CH— COAr Br' Br“ CHBr Br

It has also been suggested that the vinoxyphosphonium

interraediate could be formed from the ketophosphonium

species by rearrangement via a four-membered cyclic phosphorane.

(RO)^?*-- CH^ (R0 )^P-CH2 (RO)^P'^ CH^

0 =C — Ar 0 — Ct-Ar 0 — C — Ar

The intermediates proposed have however been hypothe­

tical species whose existence was inferred in all cases from kinetic or other studies.

18 ■xr.. V..

CHAPTER II RESULTS AND DISCUSSION

11.1 The Michaelis-Arbuzov Reaction of Triaryl Phos­ phites ...... 20

11.1.1 Mechanism of Aryl Halide Formation ...... 20

11.1.2 Disproportionation of Michaelis-Arbuzov Intermediates ...... ¿-4 1. Thermal Decomposition (in the absence of advent) of Methyltriphenoxyphosphonium Halides 2? 2 . Thermal Disproportionation (in the absence of solvent) of Dime thy Idiphenoxyphosphoniuiri Bromide ...... JD Conclusion ...... _

3. Attempted Deoxygenation of Phosphonate by Triphenyl Phosphite ......

4. Thermal Decomposition in Deuterochloroform of Methyltriphenoxyphosphonium Halides and Methyltri-o-tolyloxyphosphonium Halides and the Detection and Isolation of New Intermediate Containing a P-O-P Linkage ...... 45

5 . Hydrolysis of Disproportionation Products , 55 6 . Some Observation on N.M.R. D a t a ...... 56

11.2 Perkow and Arbuzov Intermediates ...... 52

11.3 Bisdimethylamino(methyl)neopentyloxyphosphonium Halides ......

Summary ......

19 INTRODUCTION

The work described in this thesis consists of the syste­

matic study and investigation of two aspects of the reac­

tions of phosphorus (III) esters with halogeno-compounds.

The first concerns some aspects of the Michaelis-Arbuzov

reaction of triaryl phosphites with alkyl halides and the

second a study of stabilised intermediates obtained in

the Arbuzov and Perkow reactions of neopentyl esters with

or-halogeno-compounds and alkyl halides.

II.1 THE MICHAELIS-ARBUZOV REACTION OF TRIARYL PHOSPHITES

Two aspects in particular have been considered:

1. The mechanism of formation of aryl halides (ArX) and the possible formation of benzyne intermediates.

2. The disproportionation of triphenoxyphosphonium intermediates. Intermediates from the reactions of triphe- nyl phosphite with methyl iodide and methyl bromide were prepared by a described method,as shown below:

(ArO)^P + Mel/Br — (ArO)^P‘^MeI"/Br'

II.1.1 MECHANISM OF ARYL HALIDE FORMATION

The mechanism of dearylation and formation of ArX in the second stage of the Arbuzov reaction, where the phosphonium

20 intermediate decomposes to ArX and phosphonate was inves­ tigated.

(ArO)^P'^Mer ArX + (ArO)2 P(0 )Me

This process is most simply regarded as occurring via

nucleophilic attack of halide ion. However, the formation

of ArX by nucleophilic attack on the benzene ring by a

halide anion in the absence of an activating group, i.e. NO^ is unusual.

Electrophilic substitutions in the benzene ring^^ are

promoted by OH, NR^, or alkyl groups etc., which will

direct incoming electrophiles predominantly to the ortho

and para positions, whereas electron-attracting substi­

tuents i.e. NO2 , CN and COOEt direct the electrophiles to the meta position.

Nucleophilic substitutions^'^ in aromatic rings occur

more readily when the electron density in the aromatic

system is diminished by the presence of one or more

strongly electron-attracting groups - most often nitro groups.

Substitution in non-activated rings may occur at

elevated temperatures but under these conditions a benzyne

mechanism frequently contributes to the reaction pathway.

To investigate this possibility, methyltri-o-tolyloxy-

phosphonium bromide (CH^CgH^O)^P+MeBr" was thermally de­

composed at 200 °G. The products of decomposition were

distilled under a high vacuum of O.O5 mraHg, and were

collected in a cold trap at -80 °C. They were identified by H n.m.r. spectroscopy and gas—liquid chromatography (g.1 .c.).

21 It was found that only o-tolyl bromide was present.

No traces of either m- or p-tolyl bromide were detected,

although the benzyne intermediate scheme predicts the

formation of the ortho and meta isomers.

Mechanism of Benzyne Intermediat es

,CH^ OPh PH, OPh I Br‘ HO-P'^Me Br‘ I H OPh OPh

HiOip^’Me I HBr Br; OPh OPh

CH-

This means that the benzyne intermediate hypothesis must be rejected.

Two mechanisms for nucleophilic substitution remain, the bimolecular 3^2 Ar mechanism as follows in which the

22 benzene ring is attacked by the nucleophilic Br" ion to

form bromobenzene and the stable phosphonate (Ph0 )2P(0 )Me

CH, OPh OPh ' q JL p+Me I — Br 0=P-Me ^ I Br" OPh I OPh

and the unimolecular Sj^1 Ar mechanism.

OPh

OPh

The latter is not common but can occur with a very good

leaving group e.g. in the case of aryldiazonium salts.

Ph-N. Ph"^ + N,

It should be noted that the positively charged quarter-

nary phosphorus atom is itself strongly electrophilic, so

that it may cause withdrawal of electrons from the PhO-P bond, causing an inherent tendency for dearylation to the phosphonate.

23 II.1.2 DISPROPOHTIONATION OF MICHAELIS-ARBUZOV INTERMEDIATES

The thermal decomposition of triaryl phosphite-alkyl ha­

lide adducts as described above is slow. Competitive pro­

cesses can therefore occur and it has now been shown that

disproportionation may be involved to a significant extent.

DISPROPORTIONATION is generally defined as a reaction in

which a single substance undergoes change to produce pro­

ducts, one of which is in a higher oxidation state and the

other in a lower oxidation state. The term is also

frequently used to describe the redistribution of ligands

attached to a particular atom.

It can happen that a substance is unstable because it

readily disproportionates, or in other case the single

substance reacts as both oxidant and reductant, i.e. part

of it is reduced and part is oxidised in the process. A

familiar example is decomposition of hydrogen peroxide which can be written as: ^ 2^2 2H2O + 0^.

Examples from the field of phosphorus chemistry are as follows :

1) Primary phosphine oxides disproportionate^^ at about room temperature to primary phosphines and the correspon­ ding phosphinic acids:

2RP(0)H, RPH^ + RPH(0)0H

2 ) The phosphinio acid RPH(0)0H also undergoes a charac­ teristic redox^^ reaction when heated alone to about 150 °C, one mole of primary phosphine being formed for every 2 moles of phosphonic acid:

24 3RPH(0)0H RPH2 + 2RP(0 )(0H )2

3) Phenyldichlorophosphine disproportionates on heating

into diphenylchlorophosphine and phosphorous trichloride:^^

2PhPCl2 ^9° °C ) ph^pci + PCI^

4) From the reactions of hydrogen halides with alkyl

diphenylphosphinites the expected phosphorus containing

product Ph^PiO)}! is not isolated as such but yields the

disproportionation products, diphenylphosphine and

diphenylphosphinic acid: 31

H Ph^POR jO n ^ - Ph^PiOH + RX ^ 0 - R X

2 Ph2 P(0 )H Ph^PH + Ph^PiOiOH)

In the above example disproportionation probably could be written, as below:

.OH H ,0H Ph2 P.\ P h 2P^ + HPPh, 0=P^ 0

Kxamples of disproportionation involving phosphonium intermediates have been described in detail by H.N. Rydon in 1956 in which the exchange of ArO and X anions in the triaryl phosphite dihalides leads to complex types of salt

(see Introduction).

The rapid exchange of ArO anions has also been shown to account for the disproportionation of aryloxyphosphoranes in ionising media such as CH.,CN.^^ 3 The only reported example of disproportionation invol-

25 ving the restribution of methyl groups on phosphorus is

that which occurs during the pyrolytic decomposition of

the methyl iodide adducts of diphenyl phenylphosphonite

and phenyl diphenylphosphinite at temperatures in excess of 200 ^0,^^

These quasiphosphonium salts are unusually stable to

heat and resist the normal Arbuzov cleavage. Exchange of

methyl and phenoxy groups is seen to have occurred in the

products (see Introduction), although the stage at which

exchange occurs is not clear. The unexpected formation of

dimethyldiphenoxyphosphonium bromide during the heating

of triphenyl phosphite with an excess of methyl bromide

appears to be due to a related process,although it has

not been investigated further. The effect of heat on

methyltriphenoxyphosphonium bromide, previously thought to

decompose directly to bromobenzene and diphenyl methane-

phosphonate, has therefore been re-examined.

The investigations carried out concentrated on the

following five main aspects:

1. Thermal decomposition (in the absence of solvent) of methyltriphenoxyphosphonium halides to dimethyldiphenoxy­ phosphonium halides and higher order products.

2. Thermal disproportionation (in the absence of sol­ vent) of isolated dimethyldiphenoxyphosphonium bromide to trimethylphenoxyphosphonium bromide.

3. Attempted deoxygenation of phosphonate by triphenyl phosphite.

4. Thermal decomposition (in CDCl^) of methyltriphenoxy-

26 phosphonium halides and the detection and isolation of a

new intermediate containing a P-O-P linkage.

5. Hydrolysis of disproportionation products.

In the first experiment the crystalline intermediate

methyltriphenoxyphosphonium bromide was heated without

solvent in a sealed n.m.r. tube at I75 °c. After

heating for 86.5 hours a stable clear liquid was formed,

which after cooling did not crystallise immediately but

remained as a clear, viscous mass. The n.m.r. (neat)

spectrum at this stage (Fig. 1 ) showed the phosphonium

bromide (PhO)3 P^MeBr", (41.6«%), the phosphonate (Ph0 )2P(0 )Me, (30.0%) and phenoxy protons.

There was also an additional signal at cf3.23 (Me doub­ let ) J PCH ^2, assigned to (PhO)2 P‘*’Me2Br” (28.3%), as a product of disproportionation.

Similar results were obtained after heating for 96 hours o *1 at 175 C. The H n.m.r. spectruin of an approximately 10%

solution in deuterochloroform (Pig. 2 ) showed all signals

as in the previous experiment, but with slightly different

chemical shifts in the solvent, as shown in Table I.

Confirmation of the disproportionation and that a new

carbon bond was formed in a product with two methyl groups attached to the phosphorous atom was obtained from the 31 P n.m.r. spectrum which showed a singlet at 0^ - 17.58 ppm due to triphenyl phosphate, (PhO)^PO, a quartet at d'+24.07 ppm, due to phosphonate, and quartet at cT +41.8 ppm, due to

27 Fig. 1 1 H n.iri.r. (Neat) spectrum of disproportionation products from (PhO)jP+MeBr‘ at I75 °c (86.5 h) in the absence of solvent.

Fig. 2 1 H n.m.r. (CDCl^) spectrum of disproportionation product; from (PhO)jP*MeBr- at I75 °C (96 h) in the absence of solvent.

28 unreactod methyltriphenoxyphosphoniuin bromide. A septet

at cT +96.17 ppm confirmed the presence of dimethyldiphen-

oxyphosphonium broiriide, and a small signal at cT+3 5 .98,

probably due to a P-O-P component (see later), was also observed.

Overall, a sequence of the following type appears to occur.

(PhO)^P'^MeBr“ (PhO)2P(0 )Me + PhBr

2(PhO)^P‘^MeBr" (PhO) p-^Me.Br" + (PhO),P+Br"

(PhO)^P'^Br" (PhO)^PO + PhBr

It now also seems likely that an unidentified ^^P n.m.r

peak (cT- 25 ppm), recorded previously during the heating

of triphenyl phosphite with certain alkyl halides,''^ was

due to the tetraphenoxyphosphonium ion,^^ formed in this way.

A further investigation of the disproportionation was

carried out to study the effect of prolonged heating at higher temperatures up to 250 °C.

The H n.m.r. spectra were recorded at different inter­ vals of time (without CDCl^ in this case) and are shown in Pig. 3 . It can be seen that the methyltriphenoxyphos- phonium bromide signals (A) disappeared after heating at

200 for 67 hours. Confirmation of the identity of the various products of disproportionation and decomposition was then obtained by n.m.r. in CDCl^ which showed triphenyl phosphate at cT-17.6 (s), diphenyl methanephos- phonate cT23»9 (quartet), dirnethyldiphenoxyphosphonium

29 Fig. 3 1 H n.m.r. spectra (without solvent) of disproportionation of methyltriphenoxyphosphonium bromide.

30 bromide c5* 96 (septet), and trimethylphenoxyphosphonium

bromide (T 102.4 (multiplet).

Further heating then caused an increase in trimethyl­

phenoxyphosphonium bromide (2), and also the formation of

additional products at 225-250 °C labelled (3) and (4), which were considered to be tetramethylphosphonium bromide

Me^P'*'Br*, and the phosphinate, Ph0P(0)Me2» formed by

Arbuzov cleavage of the salt (1).

These three salts are clearly much more thermally

stable than methyltriphenoxyphosphonium bromide, a fact which can be attributed to the replacement of electron- attracting phenoxy groups by electron-releasing methyl groups.

-e— GH,

To follow the course of reaction in more detail some

H n.m.r. tubes were sealed with (PhO)^P'^MeBr", and then heated at 175 °C from 2 to 8 days, and one for an addi­ tional period at 200 (24 hours at 175 °C and 96 hours at 200 °C). Each tube was opened in a atmosphere, deuterochloroform was added, and the H n.m.r. spectra recorded.

The collected H n.m.r. spectra (Fig. 4) show a picture of reaction in which the proportion of the phosphonium salt, (PhO)^P'^MeBr“ (A), fell to 6 7 ,8% after two days of heating and gradually, after 8 days, to 28, whilst the disproportionation product, (Ph0 )2 P^Me2Br"’ (1 ), and the phosphonate (Ph0)2?(0)Me (B), correspondingly increased

31 Fig. 4 -j H n.m.r. spectra (CDGl^) of thermal decomposition without solvent of methylti'iphenoxyphosphonium bromide (A), to di­ me thyldiphenoxyphosphoniuia bromide (1), to triiriethylphe- noxyphosphonium bromide (2) and to diphenyl methanephos- phonate (B). in concentration. The collected results of the above

experiments are shown in Table I.

It can also be seen that after five days of heating a

further disproportionation product, trimethylphenoxyphos- phonium bromide, PhOP'^’Me^Br" (2), began to appear. This

product reached a level of 6.3% by the time the original phosphonium salt had completely decomposed (one day at

175 and four days at 200 °C). The composition at this

stage was as follows:

(PhO)^P'^MeBr" (Ph0)2P‘^Me2Br“ PhOP‘*’Me,Br" (Ph0)2P(0)Me

0 % 21.7% 6.5% 7 2 .0%

It can also be seen (Fig. 5) that methyltriphenoxyphos- phonium bromide heated (without solvent) at a higher tempe­ rature up to 260 °C for 36 hours disproportionated to three phosphonium intermediates: dimethyldiphenoxyphosphonium

Disproportionation of (PhO)^P^MeBr” at 260 °G (Fig. 5) bromide (PhO)2 P^Me2Br~ (1.39^), trimethylphenoxyphosphonium bromide PhOP'*’Me^Br” (6 ,7%), and tetramethylphosphonium bro­ mide Me^P’^Br" (1.496), and to two phosphonates: diphenyl methanephosphonate (PhO)2P(0 )Me (8596) and phenyl dimethyl- phosphinate Ph0P(0)Me2 (6.0%) (see Table I).

Methyltriphenoxyphosphonium iodide similarly dispropor- tionated to dimethyldiphenoxyphosphoniura iodide (11.3%), and to trirnethylphenoxyphosphonium iodide (1 4 .8%)

(Fig. 6), whilst undecomposed methyltriphenoxyphosphonium iodide (3 .9%) remained (Table II).

"I H n.m.r. spectrum in CDCl^ of methyltriphenoxyphosphonium iodide after the solid had^been heated at 210 for two days. It had decomposed completely to dimethyldiphenoxy- phosphonium iodide, some of which had decomposed to trimethylphenoxyphosphonium iodide. (Fig. 6)

Methyltri-o-tolyloxyphosphonium bromide heated at 200 °C in absence of solvent for 38.5 hours showed disproportiona­ tion to dimethyldi-o-tolyloxyphosphonium bromide and to trimethyl-o-tolyloxyphosphonium bromide the results being

34 31 1 confirmed by P n.m.r. and in some cases by n.m.r. spectra for hydrolysed products (Table III).

2. Thermal disproportionation (in the absence of solvent) of dimethyldiphenoxyphosphonium bromide

In order to gain further information on the dispropor­ tionation reactions the first disproportionation product, dimethyldiphenoxyphosphonium bromide, which had been isolated previously as a crystalline solid was heated in a sealed H n.m.r. tube at its melting point (210 °C) for 190 hours. The H n.m.r. spectrum (Fig. 7) in deuterochloroform then showed signals of the starting material, dimethyldiphenoxyphosphonium bromide (36.5%), of trimethylphenoxyphosphonium bromide (35.3%) as a product of disproportionation, and diphenyl methanephos- phonate (28.2%). The P n.m.r. spectrum (Fig. 8) in deuterochloroform showed signals at cT-17.7 due to triphenyl phosphate, cT23.9 (quartet) due to diphenyl methylphospho- nate, (5^96.0 (septet) due to unreacted dimethyldiphenoxy­ phosphonium bromide and cT 102.2 (decet) due to trimethyl­ phenoxyphosphonium bromide. A very weak signal at 125.1 was probably due to triphenyl phosphite, formed by dissociation of the triphenyl phosphite complex.

The most interesting observation is that none of the expected Arbuzov product which would be derived directly from the starting material was obtained, but only those, i.e. (PhO)^PO, and (Ph0 )2P(0 )Me, which would be obtained after disproportionation to yield phosphonium ions contai­ ning three of four phenoxy groups. The other disproportio-

35

"^1 ■j H n.m.r. (CDCl^) spectrura of disproportionation products from (PhO)2P’^Me2Br- at 210 °C. (Pig. 7)

31 P n.m.r. (CDCl^) spectrum of disproportionation products from (Ph02P'^Me2Br” at 210 °C. (Fig. 8) nation product, PhOP''Me^Br” , did not undergo decomposi­

tion under tlie conditions of the experiment.

A similar experiment in which dimethyldiphenoxyphos-

phonium bromide was strongly heated up to 290 for 46

hours showed further disproportionation to trimethyl-

phenoxyphosphonium bromide (24.0%), and to tetramethyl-

phosphonium bromide (Me^P'^Br“ ) (12.5%), cf2.12 (Me, d,

'^PCH Hz), and also gave the Arbuzov products diphenyl

methanephosphonate (48.5%), and phenyl dimethylphosphinate

Ph0P(0)Me2 (1 4 .9%), cT 1.78 (Me, d, J 14.4 Hz) (Fig. 9). 31 ^ The P n.m.r. spectrum confirmed the products of dispro­

portionation with signals at cT25.9 (quartet) due to

(Ph0)2P(0)Me, 22.6 (small) due to Me^P'*’Br", 60.3 (multi- 4 plet) due to Ph0P(0)Me2, 101«9 (multiplet) due to

PhOP Me^Br , whilst the n.m.r. spectra showed

doublets fc methyl group attached to phospho follows.

Compounds JpQ (Hz

PhOP'^Me-Br" 3 13.3 64.6 M e .P^Br~ 10.6 4 55.5 (Ph0)2P(0)Me 11.5 144.7

Prom the disproportionation products of dimethyldiphen-

oxyphosphonium bromide, two components were separated,

which the H n.m.r. spectrum (DMSO) showed to be tetra-

methylphosphonium bromide, and trimethylphosphine oxide

(Fig. 10).

37 Fig. 9

H n.m.r. spectrum in CDCl^ of thermal decomposition of (PhO)2P‘^Me2Br~ at 190-290 °C

Fig. 10

H n.m.r. spectrum in DMSO of Me^P'^Br” and Me,P0 separated from reaction mixture

38 Triniethy 1 phosphine oxide was also obtained by hydro­

lysis of trimethylphenoxyphosphonium bromide and trimethyl-

phenoxyphosphonium iodide, and was identified by its 15 1 characteristic ^C- H coupling constant, 128.8 (D^O) 43 1 (Lit.^-" 129.0 in D^O), and its ' H n.rn.r. spectruia in DMSO

(cT1.50, 13.2 Hz) (Lit."^^ I.39, 13.2 Hz),

^H n.m.r. (DMSO) and (D^O) spectra of isolated trimethylphosphine oxide (Fig. 11).

^^C-Jh , J = 128.8 Hz JpCH = ^5.2 Hz

Conclusions

The systematic investigation of the disproportionation of methyltriphenoxyphosphonium bromide described above confirmed that this compound gives dimethyldiphenoxyphos- phonium bromide (PhO^P'^Me^Br“, when heated at high tempe­ rature in a sealed tube.

A similar result was also obtained with the methyltri-

39 phenoxyphosphonium iodide, q s shown:

2(PhO)^P‘^MeBr' (PhO)2P'’'Me2Br" + (PhO)^P'*'Br'

2(PhO)^P'^Mel" (PhO)2P'^Me2l’ + (PhO)^p-'r

In this way the two phosphonium intermediates

(PhO)2 P Me2Br and (PhO)2 P Me2l were isolated as pure crystalline solids. It was also found that the dimethyl- diphenoxyphosphonium halides disproportionated to give the trimethylphenoxyphosphonium salts on strong heating, both the bromide and the iodide being isolated in small amounts as crystalline solids from their reaction mixtures.

2(PhO)2?'^Me2Br PhOP'^Me^Br" + (PhO)^P‘^MeBr‘

2(PhO)2P'^Me2r PhOP'*'Me-I” + (PhO),P‘*'MeI"

Therefore the overall reactions of disproportionation which are proposed are as in the equations below:

(PhO)^P MeBr ^> (PhO)2P^Me2Br main product (2)

(PhO)2P(0)Me + PhBr 175 °C 3(PhO)^P'^MeBr' (PhO)^P'^Br" (PhO)^PO + PhBr

(PhO)2P^Me2Br" main product (2)

210 ^ G (PhO)2 P'^Me2 Br“ ------^ (PhO)P'^Me^Br" main product (3)

40 1^' * ’^*^j*^**

phenoxyphosphoniurri iodide, os shown:

2(PhO)^P‘*‘MeBr' (PhO)2P'‘’Me2Br" + (PhO)*P'^Br‘

2(PhO)^P‘^Mel" (PhO)2P'^Me2l' + (pho^p-^r

In this way the two phosphonium intermediates

(PhO)2 P Me2Br and (PhO)2 P Me2 l were isolated as pure

crystalline solids. It was also found that the dimethyl-

diphenoxyphosphonium halides disproportionated to give the

trimethylphenoxyphosphonium salts on strong heating, both the bromide and the iodide being isolated in small amounts as crystalline solids from their reaction mixtures.

2(PhO)2P'^Me2Br PhOP'^’Me^Br" + (PhO)^P‘‘’MeBr'

2(PhO)2P"^Me2l" PhOP'^Me^I" + (PhO),P'^MeI“

Therefore the overall reactions of disproportionation which are proposed are as in the equations below:

(PhO)^P MeBr ^y (PhO)2 P^Me2Br main product (2)

(PhO)2P(0)Me + PhBr 175 °C 5(PhO),P'^MeBr' (PhO)^P'^Br" (PhO)^PO + PhBr

(PhO)2P^Me2Br” main product (2)

210 ^ G (PhO)2 P^Me2Br * > (PhO)P"^Me^Br” main product (3)

40 /- ^y-vr ••

^ 2 (PhO)P'*’Me^^Br“ main product (5) 4(PhO)2P^M02Br' 2(PhO)^P'*'MeBr" {mo)^?{0)Ke

(PhO)^P'^Br" . (PhO),PO + PhBr 2(PhO)^P‘^MeBr (PhO)2P‘^Me2Br* involved in further reaction

All the above products of decomposition were identified 31 and determined by P and 1 H n.m.r. spectroscopy, and some 1 3 by the n.m.r. The exception was (PhO)^P''’Br“ which is probably not stable at the high temperatures used,

although the decomposition products (PhO)^P(O) and PhBr were found.

To explain the mechanism of disproportionation it is im- portant to compare the results with those of Rydon in 28 1956, who showed that (ArO)^PCl was obtained on recrys­ tallisation of (ArO)-PCl^. 3 2

2(ArO),PCl^ (ArO)^PCl + (ArO)2 PCl^ j 2

Although the mechanism of this process has not been investigated in detail, it is reasonable to suppose that the transfer of phenoxide and halide anions occurs by a series of reactions as shown:

ArO ArO— P Cl Cl ^— OAr' ArO ^ Cl

41 A r O ^ ArO -OAr ArO>^ A r O ~ P'^Cl OAr' ArO— P'^OAr Cl' ArO Pj 'Q '^ OAr ArO

Quite recently this interpretation was confirmed by 36 I.S. Sigal in 1979» who investigated the disproportio­

nation of aryloxyphosphoranes, in which a rapid redistri-

bution of aryloxy groups between all possible phosphora-

nes and aryloxyphosphonium intermediates were observed.

The only previous experiment was reported by Nesterov,

who showed that disproportionation of methyl and phenoxy

groups occurred in the thermal decomposition of certain

quasiphosphonium salts as follows: 24

1) MePPh(OPh)^! Phl + (Ph0)^P(0)Ph + PhPMe^I

flame 2) MePPh2(0Ph)I Phi + Ph2?(0)(0Ph) + Ph^PMe^I

The intermediate stages of the process were not studied although it is likely that a sequence such as the follo­ wing occurred:

o^„ (PhO),P'^PhI‘^ y (PhO)2P(0)Ph 1) 2Ph(PhO)2P'^MeI + Phi Ph(PhO)P'^Me2r

Then 2 mol of Ph(PhO)P‘^Me2l' Ph(PhO)2PMeI + PhPMe^I

2) 2Ph2(PhO)PMeI Ph(PhO)2P‘^PhI" + Hi2PMe2l

Ph(PhO)2P‘^PhI' Ph2(Ph0)P=0 + Phi

42 The disproportionation of phenoxy groups and methyl groups is however leas easy to understand in terms of

anion exchange.

A possible mechanism which may be proposed for the disproportionation of methyltriphenoxyphosphonium bromide

(as shown below) involves ligand exchange^ between bromide and phenoxy, the transfer of Br'*’ from the

so-formed bromophosphonium intermediate to a molecule of phosphorus III ester (formed by simple dissociation),

and further reaction of the newly formed phosphorus (III)

ester with methyl halide, as shown:

(PhO)^P'*’MeBr“ (PhO)^P + MeBr

PhO^ PhO,^ ^Me PhO. Me PhO— P'*‘-Me_ ^=25 PhO — P^ + PhO' PhO^ PhO^ Br PhO' Br

PhO^ ^ Me . OPh PhO. P h O x ^ ^P'^ +:P— OPh PMe + PhO— P^Br PhO ^ ^ Br ^ OPh PhO PhO

P h O ^ ^ M e P-Me + MeBr Pt Br- PhO' Me

P h O ^ P h O v ^ ^ P h O PhO— PiBr + PhO > P^ Br PhO"^ P h O ^ PhO

(PhO)^P'^Br" ---» (Ph0)^P=0 + PhBr

43 It is also possible that the diphenyl laethylphoophonite

may be converted to dimethyldiphenoxyphosphonium bromide

by direct reinoval of itiebhyl from another molecule of the

starting material, rather than by reaction with methyl

bromide.

PhO> OPh PhO .OPh OPh --- > PhO— P^ Me + P — - OPh Me' OPh M e ^ ^ OPh

The transfer of methyl from quasiphosphonium salts to

phosphorus (III) esters in this way has been reported by

other workers. 57

3. Attempted deoxygenation of phosphonate by triphenyl phosphite

An alternative possible route for the formation of the

tervalent ester, diphenyl methylphosphonite, was thought

to be the deoxygenation of the normal Arbuzov product

(the phosphonate) by triaryl phosphite as follows:

(PhO),P + O=P(0Ph),Me high temp.^ (ph0),P=0 + (PhO), PMe

An example of deoxygenation of phosphine oxides by

triphenyl phosphite has been claimed to give the corres- 58 ponding pho sphine s.

To investigate this possibility, triphenyl phosphite and diphenyl methylphosphonate (1:1 mol. ratio) were I heated in sealed H n.m.r. tubes, the temperature being gradually increased from 60 to 210 °C in the time

44 interval 10-24 hours, and then for up to 4 weeks at

210 °C, but no change was observed.

The tube was then opened and 1 mol. equiv. of Mel was

added and the tube reheated. The only product obtained

was methyltriphenoxyphosphonium iodide, which was identi­

fied by hydrolysis to the phosphonate (Ph0 )2 P(0 )Me.

The negative results obtained for deoxygenation,

support the mechanism of disproportionation which has

already been described.

4. Thermal decomposition in deuterochloroform of methyltri- phenoxyphosphonium halides and methyltri-o-tolyloxyphos- phonium halides and the detection and isolation of new intermediate containing a P-O-P linkage

New types of intermediate were observed during the

thermal decomposition of methyltriphenoxyphosphonium

bromide and iodide, and of methyltri-o-tolyloxyphosphonium

bromide and iodide in chloroform or deuterochloroform. A The experiments were carried out in sealed H n.m.r.

tubes which were heated at different temperatures (125 °C,

175 °C, 200 °C) and for various length of time (from 0.5 up to 112 hours).

After heating methyltriphenoxyphosphonium bromide for

15 hours in deuterochloroform at 125 or 175 °C, unexpected signals appeared at cf 2.15 (two doublets,

JpCH ^4-.4 Hz and £a. 2 Hz) (Fig. 12). These signals appeared first, before those for the phosphonate at cT 1.75 (Me doub­ let, JpQpj 17 »7 Hz), and reached maximum height after 30 to

60 hours.

45 In dilute solution they reached 45% maxiraum, and in

concentrated solution, 20 or 28%. Then if the tubes were

heated for a longer period the two doublets decreased in

intensity and finally disappeared, whereas the phospho­

nate signals grew taller.

Fig. 12 1 H n.rri.r. spectrum of thermal decomposition of methyltri- phenoxyphosphonium bromide in deuterochloroform at 125

Similar results were obtained in all the above experi­

ments which gave the same doublet of doublets at (T2.15

ppm varying in intensity from 16 to 45% maximum. Results

for methyltriphenoxyphosphonium bromide are shown in

Table XII.

The role of CHGl^ or CDGl^ is thought to be simply that

of a diluent or suitable medium in which this reaction can occur, since the same intermediate has also been detected

(although in trace amounts only) on heating the phospho­ nium salts alone.

46 Further information on the possible nature of the intermediate from methyltriphenoxyphosphonium bromide was 31 obtained from the P n.rri.r. spectrum which suggested the presence of P-O-P linkage. The proton decoupled spectrum

(Fig. 13) confirmed the presence of two doublets (cT 77.0 and cT 34.0 ppm) each with J values of about 40 Hz which is in the expected range for P-O-P coupling. The proton coupled spectrum (Fig. 14) further showed that the phospho- rus species which gave the signal at 77 ppm had a methyl group attached (appearing as a doublet of quartets) whilst the other, at 34 Ppm, did not. Other signals present were at 41.8 (quartet) due to methyltriphenoxyphosphonium bro­ mide, 23.9 (quartet) due to phosphonate, and 96.1 (multi- plet) due to dimethyldiphenoxyphosphonium bromide (small).

A further weak signal at cT 33.9, overlapping the high field phosphorus doublet, was also present. 31 The P n.m.r. spectrum (Fig. 15) for the P-O-P inter­ mediate derived from methyltri-o-tolyloxyphosphonium bromide is similar. It shows that the two phosphorus atoms appear as doublets: at cT 32.3 (Me-P-O-P, d, JpQp 32.4 Hz) and cT 77.0 (Me-P-0-P. d, JpQp 32.4 Hz and JpQp 16.0 Hz, doublet of quartets), showing that the methyl group (CH^) is attached to the second phosphorus only.

The structure of the new intermediate is not complete­ ly clear but it is reasonable to suppose that it contains the grouping CH^-P-O-P. The appearance of a doublet of doublets in the H n.m.r. spectrum of the phenoxy compound centred at cT2.15 ppni ('^pcH ^ with

47 Fig. 13 31 P n.rn.r, proton decoupled spectrum of thermal decompo­ sition of (PhO)^P'^MeBr‘” in deuterochloroform at 125 °C

Fig. 14 31 P n.m.r. protons coupled spectrum of thermal decompo­ sition of (PhO)^P‘’’MeBr~ in deuterochlorof orm at 125 °C

(PhO)^P‘^MeBr

MeP-O-P

MeP-O-P

(PhO)2P(0)Me

y

(T76.83 cT41-75 (T;5.96 ^ 2?.66

48 Fig. 15

31 P n.m.r. (CDCl^) spectra of MeP-O-P intermediate isola­ ted from thermal decomposition of methyltri-o-tolyloxy- phosphonium bromide.

what appears to be coupling also to the more remote phosphorus atom (J C£. 2 Hz) is in accord with this interpretation. The absence of a methyl group on the second phosphorus suggests a structure such as:

49 OPh OPh OPh 1 1 1 -P"^— 0— pi-OPh or CH,-pl- 1 1 5 1 OPh OPh OPh

which could be for:aed by reaction of triphenyl phosphate

(a disproportionation product - see earlier) with methyl- triphenoxyphosphonium bromide as follows:

PhO CH. OPh I \ / OPh' (Ph0)^P=0' (PhO), P"^-0— P'^— . CH, / 5 I 5 PhO OPh OPH

0 Br (PhO)2P— 0- P"^— CH- (PhBr) OPh

The formation and disappearance of this intermediate is shown in Table XII. -I The H n.m.r. spectrum (Fig. 16) of the P-O-P inter­ mediate isolated from the thermal decomposition of methyltri-o-tolyloxyphosphonium bromide showed two similar doublets at cT 2.38 (JpQj^ 13*9 Hz, JpQp^j^ 1.9 Hz), and additionally two tall signals in ratio 3:2 for the o-tolyl methyl groups attached to the two different phosphorus atoms. These signals appeared as doublets at (T 2.02 (JpoccCH cT2.18 (JpoccCH

0.9 Hz, 6h ) respectively.

The evidence therefore is that the intermediate has

50 p - 00 MD C-- • • • • • 1 • 0) 0 cr\ MD cr> O LA s T“ CM T" M -ö <—*N f t o B tp» 00 LA MD LA 1 • • 00 T~ T— CTN o LA LA C- CJ tpv tA T" CM k J- CM •H 0 0 f t 1 p - LA CM CT\ MD CM CM • • • • • • • f t o o 1 1 r ! 1 0 tA 00 00 O CA 00 f t rH tPv tA CM CM T“ k J- CM 0 0 o s B f t '— !> • (T> CT\ tA CM T~ Kt- CM O MD CM MD MD LA (A LA 0 0 C\J tA CM CM T~ tA k J- T— CM o tí 0 0 CM i n O O LA (M I > 0 1 1 • f t u 00 CM MD O k J- LA CQ v o lA VO P- 00 CM VÛ LP^ f t 0 N 0 S m 4- rp\ CM O MD Kt- MD tA ft 0 1 1 • f t 0 t P \ H MD LA tA CM CM D- O MD "sl" MD P- 00 CM MD O B ü 0 0 f t T“ T“ 00 CA MD MD O Ti n m lA CM lA tA LA c- MD CM f t p- P- P- 00 Kh P- A- cd o tP» o ♦♦ 0 0 0 rPi 0 rH LA _tí o o O O LA • O LA LA o -p • • # • CM • • 1 • o 0 o lA 00 k J- T“ MD LA CM 0 § VD CT»

51 two o-tolyloxy groups on one phosphorus atom and three on

the other, suggesting the structure shown.

OC.H,•CH, I 5 4 3 I 6 4 3 CH,— -n+______0 ----- P'^-OC^H. • CH j 6 4 OC^H.•CH, OC^H,•CH, 6 4 3 6 4 3

n.m.r. (GDCl^) spectrura of MeP-O-P intermediate isola­ ted from thermal decomposition of methyltri-o-tolyloxy- phosphonium bromide (Fig. 16).

The above MeP-0-P intermediate which was isolated in

a very small amount was found to be stable in CDCl^

in a sealed tube, but was too unstable to be isolated

and stored as a solid even under a vacuum.

52 An interesting feature of the thermal decomposition of methyltri-o-tolyloxyphosphoniura bromide in deuterochloro- form was that the solution became blue when hot, and slowly changed to green and then to light brown or colour­ less after cooling. (See Table VII). The blue colour was observed when hot only as long as the two doublet signals were present. After these signals had disappeared the colour was changed to brown. Another specific observation was the change in height of two aromatic signals (back reaction) (T 7.33 (s) phosphonium, and 0" 'J,^2 (s) phospho­ nate. On cooling and changing from blue to green the phosphonium signals increased from 29*7% to 42.5% after

1.25 hours whereas the phosphonate signal decreased in height from 70»3% to 57*5% (Pig* 17)*

The blue colour is considered to be a specific indica­ 52 tor for phenoxy radicals.

The possible role of such radicals in the Arbuzov reac­

tion described above is however uncertain.

53 m

'd

o

54 5 , Hydrolysis of diaproportionation products

The products of thermal decomposition of dimethyldiphe- noxyphosphonium bromide (Fig. 9) were hydrolysed by atmospheric moisture to give three phosphonates: Me^PO,

Ph0P(0 )Me2 > and (PhO)2P(0 )Me (Fig. 18). The presence of the last, which was unexpected, supports the mechanism of disproportionation;

2(PhO)2P'^Me2Br‘ PhOP'^Me^Br“ + (PhO)^P'^MeBr'

(PhO^P'^MeBr (PhO)2P(0)Me

1H n.m.r. (CDCl^) spectrum of hydrolysis of disproportio­ nation products for (PhO)<5P'^Me^Br” at 290 (Fig. 18).

The formation of several different hydrolysis products when the tube of the P-O-P intermediates was opened to the atmospheric moisture, suggests that the system is

55

é complicated. The results of hydrolysis are summarised in

Tables IX and X. 51 Further studies based on P n.m.r. spectroscopy showed two stages of hydrolysis. The first stage of hydrolysis showed main, but unidentified signals at

J £a. 20.0 Hz) and d* 84.4 (m).

The second stage of hydrolysis showed a change to new unidentified signals at cTl1.7 (triplet, J 25.3 Hz) and cf 47.7 (m) probably two overlapping quartets.

6. Some observations on n.m.r. data

The chemical shift of the Me^PO peak in the decoupled 51 P n.m.r. spectrum was found to vary depending upon which solvent was used and upon the pH of the solution:

Solvent CDCl^ CDCl., CDGl^ DMSO 3 ^ 2 ^

^^P found 76.6 40.0 57.3 46.4 53.7 in reaction at pH 4 mixture lit.55 — 59.0 - - - iit.54 36.2

Other workers have found similar effect. 55 Thus the

31 P chemical shifts of Ph^PO in some different solvents 3 at concentration of 1:20 mole ratio were as follows:

56 Solvent (ppm)

1,4-Dioxane 24.8

m-Cresol 36.4

Trifluoroacetic acid 48.1

Sulphuric acid 59.8

The difference between the shifts obtained in highly acidic solvents can be interpreted raostly in terms of a large change in the position of the equilibrium:

Ph,P=0 -f HA Ph,P=0---HA Ph,P‘^-OH A" 3 3 3

In such solvents the predominant effect responsible for the downfield trend with increasing acidity appears to be that of an increase in the number and strength of solvent-to-solute hydrogen bonds as represented by the hydrogen bonded complex in the above equilibrium.

There is ample evidence in the literature that such interactions are important in systems containing triphe- nylphosphone oxide and alcohol, phenols, carboxylic acids, or chloroform.

The coupling constants for Arbuzov intermedia­ tes and products show large differences, depending on the number of methyl groups attached to phosphorus. With increasing number of methyl groups the value of decreases significantly as shown and this can be a valuable aid to their identification.

57

T í

ü

O CÖ

0

0

0

T í 0 -P 0$ m

CÖ 4-J CÖ T í

61 II.2 PERKOW AND ARBUZOV INTERMEDIATES

The reactions of trialkyl phosphites with a-halogeno- carbonyl compounds may yield two different products, either the ketophosphonate (Arbuzov product) or the vinyl phosphate (Perkow product), depending on the halogen involved and the particular reaction conditions.

There have been several proposed mechanisms for these reactions.

1. Chopard et al. suggest that the Perkow product is

formed by initial attack at the carbonyl carbon atom,

followed by migration of the phosphorus atom from the

carbon to oxygen atoms.

2. Koziara et al. propose the formation of a halogeno-

phosphonium-enolate ion pair, which is common to both

the Arbuzov and Perkow intermediates and their products.

3. Marquading and Ramirez also suggest a common interme­

diate. The ketophosphonium intermediate gives rise to

the Arbuzov product directly and to the Perkow product

by rearrangement via a four-membered cyclic phosphorane.

4. Petnehazy et al. also favour the idea of a common inter­ 60 mediate after work on a-halogenoacetophenones.

However up to now all the evidence for and against the various postulated mechanisms has been of an indirect nature particularly from kinetic studies, because no inter­ mediates have been isolated nor identified, and no direct

62 evidence has been obtained.

In my work I have isolated and identified some of the

intermediates during these reactions, and have therefore produced the first direct evidence for and against the postulated mechanisms.

The present work has now shown that identifiable alkyl-

oxyphosphonium intermediates are obtainable by the reaction

of neopentyl esters of phosphorus (III) acids with phenacyl halides. Phenacyl bromide gave the ketophosphonium salts.

I. ROP + PhCOCH^X RO-Ij— CH^COPh X" \ 2

e.g. reaction with trineopentyl phosphite in acetone at

room temperature for 2 hours allowed the isolation of the

Arbuzov intermediate (R0)^P'*’CH2C0PhBr’ which was washed

with ether and obtained as a crystalline solid, m.p.

88 °C, in 21% yield. Similarly, the reaction of dineo­

pentyl phenylphosphonite with phenacyl bromide in ether

at room temperature gave the Arbuzov intermediate

(RO)2 P**'(Ph)CH2 COPhBr" m.p. 120 - 122 °C, in 45.8?é yield,

and the reaction of neopentyl diphenylphosphinite with

phenacyl bromide in chloroform at room temperature pro­

duced (RO)P‘*’Ph2 CH2 COPhBr" m.p. 155 - 136 °C, in 71*4%

yield.

It should, however, be noted that some of the phenacyl

intermediates did not precipitate well with ether and for

this reason they were precipitated by addition of an

excess of chloroform.

63 The last pi’oduct was the most stable and was suffi­ ciently resistant to hydrolysis to allow an X-ray

structure determination to be cari’iod out in the open laboratory atmosphere without the need for special precautions.

Fig. 19

X-ray crystal structure of Ph2 P^(0R)CH2CJ0Ph Br (R = neopentyl)

64 Important Bond Lengths, and Bond Angles

Bond Lengths (A)

P - C(1 ) 1.798(11 )

1.792(11 )

1.786(12)

1.575(8)

Bond Angles (°)

C(2P) - P - C(3P) 111.8(6)

C(3P) - p - 0(1 ) 109.5(5)

C(5P) - p - C(1 ) 112.3(6)

C(2P) - P - C(1) 113.0(5)

C(1 ) - P - 0(1 ) 107.8(5)

C(2P) - P - 0(1 ) 102.0(5)

The X-ray diffraction data confirm the phosphonium structure, with bond angles at phosphorus approximating to a tetrahedral arrangement, and with a ?■*■....Br" o interatomic distance of 4.229 A. Some double-bond character between phosphorus and oxygen is indicated by the bond length of 1.573 A. The fact that the stability of these intermediates decreases as the number of alkoxy groups increases, suggests however that the inductive effect (-1) of oxygen is more important than its mesomeric (+M) effect.

65 When the reaction of (RO)^P, (R0)2PPh and of R0PPh2 with phenacyl chloride were carried out in CHCl^ at room temperature no Arbuzov intermediates or products were detected by n.m.r. spectroscopy. With (RO)^P and

(R0)2PPh> only the Perkow products were detectable, but

(R0)PPh2 gave both the Perkow intermediate and the vinyl phosphinate as detectable species.

RO^ Ph 0 Ph II / RO-P + 0=C RO-P-O-C^ RCl / 1 \ RO CH2CI OR CH,

R O ^ Ph 0 Ph H / Ph-P + 0=C Ph- P-O-C^ RCl 1 V CH2CI OR CH,

RO, Ph OR ^Ph 0 Ph « / Ph-P + 0=C — ->. Ph-P'^-O-C. Cl' Ph-P-O-C,. + RCl I V • P h / CH2CI Ph CH2 Ph CH,

35% 45% (Intermediate) (Product)

Table XIV shows the various Arbuzov and Perkow interme­

diates that were detectable in solution during the course

of this work. The reaction of phenacyl chloride with (RO)PPh2 was

therefore repeated at -5 to 0 °C, when the vinyloxyphos-

66 in-:’?**'

6 7 phonium chloride (i. e. The Perkow intermediate) was isolated and purified. Its decomposition at 33 °C in

CDCIt was followed by H n.m.r. spectroscopy, the kinetics 3 being found to be first order with a half life of approximately 40 min. (Fig. 20).

The decomposition of neopentyloxydiphenyl-1-phenylvinyloxy- phosphonium chloride in CDGl^to 1-phenylvinyl diphenylphos- phinate (Fig. 20).

83.42%

*1 A fter 60 min. 16.5896 jl

t L . . t I _ V — f -AJ. .1 . 1 : I ; i ;■ I • • • I ! ■ ’ ‘ : 75.3|i% ! • m ■ ■ m m I ■ ' I I I ■ j After 35 min.

r

I ' ' i I

i , i ‘ j ' ■* - • -1 t f f i . 65.1696: . ! ' i ! I ! ! I After 15 min.

Ph2P‘*^(0R)0C( tCH2)Pb Cl" 1 : j zCCH2Bt ! : . : Ph2P(0)OC(:CH2)Ph • 1 i i : ^-:.l i i :j I M *' I‘ ■ i' I i ■ . Ii 45.6896 ;54-?2?6l I I -1 - I * 1 ' - I

68 Perkow intermediates were characterised by the multiplet in the n.m.r. spectrum at cT 5.41-5.54 ppm due to vinyl protons and which was gradually replaced by a corresponding multiplet at cT 5.05-5.25 Ppm assigned to the vinyl protons of the product (Fig. 20). Neopentyl chloride was also formed.

The Arbuzov intermediates were considerably more stable than the Perkow intermediates but they decomposed on heating in chloroform or other solvent. It is important to note that in doing this they yielded Arbuzov products exclusively, thus showing that rearrangement to the Perkow intermediates

(as suggested elsewhere^*^) can be excluded under the condi­ tions of these experiments. Dealkylation of all intermedia­ tes gave neopentyl halides as the only halide products,

showing that the reactions occurred by an Sj^2 - type clea­ vage of the R-0 bond.

P — GH^COPh RBr 0 = P C H 2 C O P h

CH, Cl'^R— 0 2^p+— 0-C RCl — C Ph '^Ph

The observed first order kinetics are in accord with

the existence of these intermediates as ion-pairs in the

organic solvent used.

The isolated Perkow and Arbuzov intermediates and

products of decomposition were identified by the

n.m.r. and ^^P n.m.r. spectroscopy. (Table XV).

69

II.3 BISDIMETHYLAMINO(METHYL)NEOPENTYLOXY- PHOSPHONIUM HALIDES

Alkyl phosphorodiamidites (R^N)2P0R undergo typical

Michaelis-Arbuzov reactions with a number of organic hali- . . n 61 des, although phosphorus-nitrogen fission may also occur.

Few examples of reactions involving simple alkyl halides have been reported and the results are not entirely

clear. Thus the ethyl esters (R = Et ; R» = Et or Pr^)

reacted exothermically with iodomethane to give unstable

oily adducts which yielded the corresponding alkanephos-

phonic diamides on standing.

(R'N)^POR + R”X -- > (R»N)^P(OR)R"X (R^N)2P(0)R" + RX

In addition, the diethylamino derivative (R = R ’ = Et)

gave diethylammonium iodide. The only other phosphonium

species reported in reactions of this type was obtained

from the phenyl ester (R = Ph, R» = Bu^) by heating with

two equivalents of iodomethane.

Some investigations of the reactions of R0P(NMe2)2 . 48 MeX (X « Cl, Br, I) were carried out by A.R. Qureshi.

Only the Mel gave a pure phosphonium salt. The bromide was

contaminated with Me^NBr and the chloride could not be

obtained. Further studies of these reactions have therefore been

carried out.

71 The reaction of neopentyl N,N,N ',N '-tetramethylphos- phorodiamidite with iodomethane gave the alkoxyphospho- nium iodide as the exclusive product.

(a) + MeX (Me2N)2P'^(0R)MeX“

The corresponding phosphonium halides (X = Br or Cl) were also the major products of reaction with bromo- methane and chloromethane but an increasing tendency to quaternization (side reaction) at nitrogen in the order I < Br Cl led to phosphorus-nitrogen fission and for­ mation of the tetra-alkylammoniiim halides which were isolated and positively identified by IR and chemical analysis. (See Table below).

Table XVI Products from the reactions of neopentyl N,N,N’,N»-tetra- methylphosphorodiamidite with halogenomethanes

X in MeX Reaction products (mole %)

(mol. equiv.) ( M e ^ N ^ Me2 NP(0 R)X Me^NX

I (1 .0) loo­ 0 0 Br (2.6) se- (88.6)- 14 3.9

Cl (2.5) 83- (30.1)- 17 10.4

^ Yields by n.m.r. analysis of product mixtures - Isolated yields.

72 The cleavage reactions involve nucleophilic displacement of trialkylamine from trivalent phosphorus by halide ion and give the alkyl phosphoramidohalidites (X = Cl or Br), which were detected in solution by 31 P and 1 H n.m.r. spectroscopy. Although the 31 P chemical shifts of the phosphoramidohalidites (R0PGlNMe2 cTp 178.44 and R0PBrNMe2 198.29) are close to those of the corresponding phos- phorodihalidites (ROPCl^ cTp 178.31 and R0PBr2 cTp 200.44) and cannot be used to distinguish these species with

certainty.

Me2N _____ (b)_____ \ (Me2N)2P0R + MeX side reaction P — OR Me,

MeX M62NP(0R)X + Me^N ( ^ Me^NX)

the possible formation of the dihalidites by further cleavage was excluded in separate experiments.

Me2NP(0R)X + MeX ROPX2 + Me^N

The proton n.m.r. spectra of these compounds are

distinctive: for Me2 NP(0 R)X cT^ 5.65 or 3.55 (JpocH

7 .2 Hz), for ROPX2 3.83 or 3.88 (JpQQ^ ^z). (See Table XVII). The absence of reaction between the phos­ phoramidohalidites and bromomethane or chloromethane is attributed to the electron-withdrawing effect of the halo- gen and to a consequent reduction in the tendency of the

73 -yi_

halidite to undergo quatornization at either phosphorus

or nitrogen.

The bisdirnethylamino(methyl)neopentyloxyphosphonium

halides were thermally stable and were not hydrolysed

appreciably at room temperature, the bromide and iodide

being well-defined crystalline solids that could be

handled in the open laboratory for considerable periods

of time without difficulty. The chloride, in contrast,

had a relatively low melting point, was difficult to

crystallise, and was highly deliquescent; nevertheless, it

underwent no detectable hydrolysis in aqueous solution

during several hours at room temperature. (Some hydrolysis

on long standing was indicated for all halides by the

development of an odour of dimethylamine during storage).

n.m.r. spectroscopy confirmed the tetracoordinate

phosphonium structure for all three compounds in solution,

the chemical shifts (ci^ ca. 60 ppm) being independent of

the halogen present. In solvents such as CDCl^ they may

be expected to exist as ion-pairs.^’ ^ Thermal decom­

position in deuterochloroform was negligibly slow at room

temperature but occurred at 100 °C to give neopentyl ha­

lides without rearrangement of the neopentyl group, indi­

cating Sjj2-type cleavage of the alkyl-oxygen bond:

^ (Me2N)2P(0)Me + RX — Me

(R = Me^CCH2)

74 The early work of Michaelis indicated the formation of diethyl ammonium iodide in addition to methylphosphonic bisdiethylamide during decomposition of bisdiethylamino-

(methyl)ethyloxyphosphonium iodide in ether but in the present studies no evidence was found for the formation of dialkylammonium salts when water was rigorously excluded.

(Et2N)2P(OEt)MeI (Et2N)2P(0)Me + EtI + Et2NH2l

An n.m.r. peak assignable to the dimethylammonium cation

(cT 2.66 ppm) appeared in certain circumstances, however,

during thermal decomposition in deuterochloroform and was

shown to be enhanced if water was deliberately added.

Under these conditions, hydrolysis of the bromide led to

the formation of small but identifiable amounts of

dimethylammonium bromide and the dimethylammonium

neopentyl methanephosphonate and bisdimethylammonium

methanephosphonate:

+ 2E^0 + (Me2N)2P(0R)Me X" ---=— » Me^NH^ •0P(0)(0R)Me + Me^NH^X '2 2 2 2 ‘

+ 3H^0 + ----— > (Me^NH^)^ -0^P(0)Me + ROH + HX (Me2N)2 P(OR)Me X' '2 2'2 2'

(R = Me,GCH«)

The n.m.r. and ^^P n.m.r. data for bisdimethylami- no(methyl)neopentyloxyphosphonium halides and their decom­

position products and starting materials are in Table XVII,

75 The relative importance of inductive and mesomeric effects of attached ligands in the stabilisation of alkoxyphosphonium salts has been discussed previously.

For alkoxy and phenoxy ligands the inductive effect of 6Æ oxygen was considered to be more important. In the examples of alkoxyphosphonium salts carrying nitrogen ligands, as in the compounds referred to here, the mesomeric (electron-donating) influence of nitrogen appears to be more important in view of the stabilising effect of the dialkylamino groups. This conclusion is supported by recent X-ray diffraction studies of the bisdimethylamino(methyl)neopentyloxyphosphonium iodide which have shown the nitrogen atoms to be planar, with phosphorus-nitrogen bond lengths corresponding to 65 significant double bond character. The P....I distance

of 4.987 A confirms the phosphonium structure for the

compound. Bond angles at phosphorus also demonstrate the

tetrahedral structure.

Bond lengths (Â)

P(1 ) - 0(1) 1 .5 4 6 (7 ) p(i) - S(1) 1.581 (10)

P(1 ) - H(2) 1.609(11 ) p(i ) - C(3) 1 .7 6 8 (1 4 )

11(1 ) - C(12) 1 .506(21 ) N(1 ) - 0(11 ) 1 .408(25)

N(2) - C(22) 1 .440(1 4 ) N(2) - 0(21 ) 1 .4 6 7(1 8 )

76 Bond Angles (*’)

N(1) - 0(1) 114.1(4) N(2) - P(1) - 0(1) 100.3(5) N(2) - N(1) 109.2(6) C(3) - P(1) - 0(1 ) 109.4(5) C(3) - N(1) 108.6(6) C(3) - P(1) - H(2) 115.5(6) C(12). P(1) 119(1 ) C(11 )- N(1 ) - P(1 ) 125(1) C(21 )- P(1) 121.6(8) C(12)- N(1 ) - C(11) 114(1) C(22)- C(21) 112(1) C(22)- N(2) - P(1 ) 119(1)

Fig. 21

X-ray crystal structure of (Me2 N)2 f(OR)Me I (R = neopentyl) co (U (H ft

<0 Cd

78 SUMMARY

Preparation and properties of guasiphosphonium intermediates

During this work 18 crystalline solid quasiphosphonium intermediates have been isolated. A ll were prepared in dry conditions by nucleophilic attack of a phosphorus

(III) ester on alkyl halide and with the use of a range of different temperatures (from below 0 to 290 °C) and reactions times (between 2 and 190 hours). Some were prepared without solvents, while others were prepared in solvents, such as acetone and chloroform. None of the phosphoniiim intermediates was soluble in ether, and it

is for this reason that the intermediates were generally precipitated from reaction mixtures by the addition of

sodium-dried ether.

A ll were washed with dry ether, dried under high

vacuum and analysed by and ^ P n.m.r. spectroscopy.

Where possible they were also assayed quantitatively for

C, H, P, and halogen.

Stabilities

The alkyloxyphosphoniura intermediates are more

resistant to hydrolysis than the aryloxyphosphonium

intermedlatea» they are, however, lees thermally stable

than the latter. A factor influencing the stability of

the phosphonium intermediates is the relative number

of RO and R* groups attached to phosphorus. Thus in the

structure, (R 0)„ P ^ R ’4_n X" (H=Ph neopentyl,

79 R'=Me or Ph| n=0 to 3 ) , the stability of the intermediate is found to increase as n decreases. That is, the repla­ cement of electron-attracting (-1) alkoxy or aryloxy groups by electron-releasing (+1) alkyl or aryl groups causes an increase in stability. The attachment of nitrogen ligands to phosphorus also causes a considerable increase in stability, possibly associated with electron donation from nitrogen to phosphorus (+M) by interaction.

Steric factors, as in the neopentyl group, are also important in providing stabilisation. The possibility of isolating reaction intermediates, whose properties and reactions can then be studied under controlled conditions, make possible a much clearer and more precise understanding of reaction mechanisms than was possible before.

80 CHAPTER III EXPERIMENTAL

Starting Materials ......

Analytical Techniques and Instrumentation

III.1 METHYLTRIARYLOXYPHOSPHONIUM HALIDES AND THEIR DECOMPOSITION PRODUCTS

Preparation of Methyltriphenoxyphosphonium Bromide.

Preparation of Diphenyl Methanephosphonate ....

Attempted Deoxygenation of Diphenyl Methanephospho­ nate by Triphenyl Phosphite ......

Preparation of Tri-o-tolyl Phosphite ......

Preparation of Methyltri-o-tolyloxyphosphonium Bromide ......

Preparation of Methyltri-o-tolyloxyphosphonium Iodide ......

Thermal Decomposition of Methyltriphenoxyphospho­ nium Bromide in the Absence of Solvent ......

Identification of the Products of Thermal Decompo­ sition of Methyltriphenoxyphosphonium Bromide in the Absence of Solvent ......

Thermal Decomposition of Methyltriphenoxyphospho­ nium Iodide in the Absence of Solvent and Identi­ fication of the Products ......

^H n.nL.r. data of Thermal Decomposition of Methyl- tri-o-tolyloxyphosphonium Bromide and Iodide in the Absence of Solvent ...... Isolation of Dimethyldiphenoxyphosphonium Bromide and Hydrolysis of Dimethyldiphenoxyphosphonium Bromide ......

81 Thermal Decomposition of Dimethyldiphenoxyphospho- nium Bromide ......

Isolation of Trimethylphenoxyphosphonium Bromide containing some Dimethyldiphenoxyphosphonium Bromide

Hydrolysis of Trimethylphenoxyphosphonium Bromide .

Isolation of Mixture of Tetramethylphosphonium Bromide containing Trimethylphosphine Oxide ....

Isolation of Trimethylphosphine Oxide ......

Isolation of Dimethyldiphenoxyphosphonium Iodide. .

Hydrolysis of Dimethyldiphenoxyphosphonium Iodide .

Thermal Decomposition of Dimethyldiphenoxyphospho­ nium Iodide ......

Isolation of Trimethylphenoxyphosphonium Iodide con­ taining traces of Tetramethylphosphonium Iodide . .

Isolation of Trimethylphosphine Oxide from the Thermal Decomposition of Methyltriphenoxyphospho- nium Iodide ......

Hydrolysis of Methyltriphenoxyphosphonium Bromide in Deuterochloroform ......

Hydrolysis of Methyltriphenoxyphosphonium Iodide in Deuterochloroform ......

Hydrolysis of Methyltri-o-tolyloxyphosphonium Bromide in Deuterochloroform and in Air ......

Hydrolysis of Methyltri-o-tolyloxyphosphonium Iodide in Deuterochloroform ......

Thermal Decomposition of Methyltriphenoxyphospho­ nium Bromide in Deuterochloroform in a Sealed •1 H n.ra.r. Tube ......

Thermal Decomposition of Methyl Triaryloxyphospho- nium Halides in Deuterochloroform under various conditions ......

Hydrolysis of P-O-P Intermediate obtained from

82 Methyltriphenoxyphosphonium Bromide in Deuterochlo- roform ......

Thermal Decomposition of Methyltriphenoxyphosphonium Iodide and Simultaneous Distillation of Products . .

Hydrolysis of the P-O-P Intermediate Formed during Thermal Decomposition of Methyltriphenoxyphosphonium Iodide in the Absence of Solvent ......

Thermal Decomposition of Solid Methyltri-o-tolyloxy- phosphonium Bromide and Distillation under High Vacuum ......

Isolation of MeP-O-P Intermediate after Thermal De­ composition of Solid Methyltri-o-tolyloxyphosphonium Bromide and Distillation of Volatile Products . . .

III.2.1 OBSERVATION OF THE COURSE OF THE REACTIONS BETWEEN PHENACYL HALIDES AND PHOSPHORUS (III) ESTERS IN DEUTEROCHLOROFORM BY 'H N.M.R. SPECTROSCOPY

(a) Trineopentyl Phosphite and Phenacyl Bromide . .

(b) Dineopentyl Phenylphosphonite and Phenacyl Bromide ......

(c) Neopentyl Diphenylphosphinite and Phenacyl Bromide ......

(d) Trineopentyl Phosphite and Phenacyl Chloride . .

(e) Dineopentyl Phenylphosphonite and Phenacyl Chloride ......

(f) Neopentyl Diphenylphosphinite and Phenacyl Chloride ......

III.2.2 PREPARATION AND ISOLATION 0F PHOSPHONIUM INTERMEDIATES

The Arbuzov Intermediates

83

Ü 1. Preparation of Trineopentyloxy(phenacyl)phospho- nium Bromide ......

2. Preparation of Dineopentyloxy(phenacyl)phenylphos- phonium Bromide . . • ......

3. Preparation of Neopentyloxy(phenacyl)diphenyl- phosphonium Bromide , ......

The Perkow Intermediate

1, Preparation of Neopentyloxydiphenyl-1-phenylvi- nyloxyphosphonium Chloride ......

III.2.3 OBSERVATION OF THE COURSES OF THE DECOMPO­ SITION OF ISOLATED PHOSPHONIUM INTERMEDIATES IN DRY CDCl^ BY ^H N.M.R. SPECTROSCOPY 1>

Decomposition of Arbuzov Intermediates

1. Decomposition of Trineopentyloxy(phenacyl)phospho-

nium Bromide in Deuterochloroform • • • •

2. Thermal Decomposition of Dineopentyloxy(phenacyl)- phenylphosphonium Bromide in Deuterochloroform

3. Thermal Decomposition of Neopentyloxy(phenacyl)- diphenylphosphonium Bromide in Deuterochloroform

Decomposition of the Perkow Intermediate

1. Decomposition of Neopentyloxy(diphenyl)-1-phenyl- vinyl oxyphosphonium Chloride in Deuterochloroform

III.2.4 PREPARATION AND ISOLATION OF ARBUZOV AND PERKOW PRODUCTS

Arbuzov Products 1. Preparation and Isolation of Neopentyl Phenacyl- (phenyl)phosphinate ......

2. Preparation of Phenacyl(diphenyl)phosphine Oxide

Perkow Products

84

M m z -*. / ' I 1. Preparation of Neopentyl 1-Phenyl vinyl Phenyl- phosphonate ......

2. Preparation of 1-Phenylvinyl Diphenylphosphi- n a t e ......

III. 3 PREPARATION AND THERliAL DECOMPOSITION OF THE INTERMEDIATES FORMED BY REACTION OF NEOPENTYL N,N,N»,N'-TETRAMETHYLPHOSPHORODIAMIDITE WITH HALOGENOMETHANES

Preparation of Neopentyl Phosphorodichloridite

Preparation of Neopentyl N,N,N’,N’-Tetramethyl- phosphorodiamidite ......

Preparation of Neopentyl Phosphorodibromidite . .

Preparation of Neopentyl N,N-Dimethylphosphoroami- dochlorodite ......

Preparation of Neopentyl N,N-Dimethylphosphoraini- dobroinidite......

Preparation of Bisdimethylamino(methyl)neopentyl- oxyphosphonium Chloride ......

Preparation of Bisdimethylamino(methyl)neopentyl- oxyphosphonium Bromide ......

Preparation of Bisdimethylamino(methyl)neopentyl- oxyphosphonium Iodide ......

Investigation of the Action of Halogenomethanes on N,N-Dimethylphosphoramidohalidites ......

(a) Cliloromethane ......

(b ) Bromomethane ......

Thermal Decomposition of Bisdimethylamino(methyl)- / neopentyloxyphosphonium Halides ......

(a) In anhydrous CDCl^ ......

(b ) In presence of Water ......

Absorbtion of Water by the Phosphonium Chloride .

Bibliography

85 STARTING MATERIALS

Alcohols

Neopentyl alcohol (2,2-Dimethyl- Aldrich Chem.Co., Ltd. propan-1-ol)

Amines

N,N-Dimethylaniline G.P.R Hopkin & Williams Ltd. TDried barium oxide, redis­ tilled 66 °C at 6 mmHg)

N ,N-Dimethylamine

Pyridine (kept over potassium hydroxide)

Halides

Methyl bromide Fine Chemicals It

Methyl chloride cylinder B.D.H.Chemicals Ltd.

Methyl iodide G.P.R Hopkin & Williams Ltd. (Dried over sodium sulphate, redistilled b.p. 43 C)

a-Bromoacetophenone Aldrich Chem.Co., Ltd

a-Chloroacetophenone

o-Cresol (redistilled 67 C Hopkin & Williams Ltd. at 6 m m H g )

Phosphorus compounds

Triphenyl phosphite (redis­ Aldrich Chem.Co., Ltd, tilled 172 C at 0.5 mmHg)

Phosphorus trichloride G.P.R. Hopkin & Williams Ltd, (redistilled b.p. 73-74 G)

Phosphorus tribromide G.P.R

86

ChlorodiphenylphocpJiine Aldrich Chern.Co., Ltd. (redistilled 120 *^0 at 1.0 mmHg)

Dichlorophenylphosphine

Solvents

AnalaR Acetone (refluxed with po­ Hopkin & Williams Ltd. tassium oermanganate, distilled b.p. 56 C)

Anhydrous ether (kept over sodium wir e )

Chloroform AnalaR (distilled grom It phosphorus pentoxide b.p. 6l C)

Petroleum spirit b.p. 30-40 *^G (kept over sodium wire) It

87 Analytical Techniques and Instrumentation

Carbon-Hydrogen-Nitrogen Analysis

These elements were determined in the Microanalytical

Laboratory of the Department using a Perkin-Elmer 240

Elemental Analyser. The samples were submitted in a small desiccator to protect them from moisture.

Halogens

Chlorine, bromine, iodine were determined by Vohlard's method.

A sample of halogen - containing compound (ca. 0.5 g) was weighed out in a sample tube, and was added to a stoppered conical flask containing 2.0 g potassium hydro­ xide in 100 ml H^O. Then the sample tube was reweighed. As soon as the sample was dissolved, 2M nitric acid (25 inl) and an excess of 0.1M silver nitrate solution were added.

The sample was then back titrated against 0.1M ammonium thiocyanate solution with ferric alum as indicator.

Phosphorus

The sainple (0.2-0.5 g) was heated with concentrated sulphuric acid (c^. 10 ml) for 10 to 20 h in a Kjeldahl flask until clear. In some cases selenium (Kjeldahl cata­ lyst) tablets were added. The contents were then cooled and concentrated nitric acid (ca« 10 ml) was added and the mixture was heated again until nitrous fumes were no longer

88 • MWS'- •»

evolved. The contents of the flask were then cooled, washed into a beaker and neutralized with 0.88 ammonia to methyl red indicator, and just acidified. Magnesia mixture

(25 ml) was added, then phosphorus was precipitated as magnesium ammonium phosphate hexahydrate by adding 0.88 ammonia solution according to the method described by

Vogel,After 24 h the precipitate was filtered on a

No. 4 sintered glass crucible and washed with dilute

ammonia solution. The precipitate was redissolved in hot

105^ hydrochloric acid and reprecipitated with magnesia mixture (2 ml) and 0.88 ammonia. After 24 h the final

precipitate was washed with dilute ammonia, rectified

spirit and anhydrous ether, dried in a desiccator and

weighed as MgNH^PO^.6H2O.

Nuclear Magnetic Resonance Spectra

n.m.r. spectra were recorded on a Perkin-Elmer

R12B spectrometer operating at 60 MHz with tetramethyl-

silane (TMS) as an internal standard.

^^P n.m.r. spectra were obtained on a Bruker WP80

spectrometer at 32.395 MHz, with 83% phosphoric acid as

an external standard.

n.m.r. spectra were obtained on the Bruker WP80

spectrometer at 20.12 MHz, with TMS as internal standard.

Infrared Spectroscopy

Spectra were recorded using neat liquids on a Perkin-

Elmer 137 KBr infraoord spectrometer. The KBr disc method

was also used.

89 Refractive Indices

These were measured on a Bellingham and Stanley Ltd.

Refractometer, thermostatted at 20 °C.

Gas Liquid Chromatography

A Perkin-Elmer 11 apparatus with flame-ionization de­ tector and nitrogen as a carrier gas was used. The glass column was 1cm x 5m with 10% PBGA on Celite (60-80 mesh).

The inlet pressure was 17 p.s.i. and the nitrogen flow- rate 20 ml/min at 115 °C.

Mass Spectroscopy

Spectra were obtained on an A.E.I. MS9 instrument operating at an electron-impact energy of 70 e.V. and with a source temperature of 180-200

X-Ray Crystallography

Determinations were performed by Dr. K. Henrick (PNL), on a Phillips PW1100 computer-controlled single crystal

X-ray diffractometer. The sample (0.5 mm cube) was sealed

into a glass tube and was loaded to the diffractometer

centre automatically.

90 HI.1 METHYLTRIARYLOXYPHOSPHOMIUM HALIDES AND THEIR DECOMPOSITION PRODUCTS

Preparation of Methyltriphenoxyphosphonium Bromide^^

Triphenyl phosphite (21.6 g, 1.0 mol. equiv.) was placed

in a thick-walled glass tube with a constriction for

sealing and corked. The tube was cooled to -80 ^C in a mixture of acetone and dry ice. An ampoule with methyl bromide was cooled first to 0 °C in an ice-bath and then

to -80 °C in the acetone-dry ice mixture. It was then

carefully opened and liquid methyl bromide (approx. 5.6 ml,

9.69 g, 1.46 mol. equiv.) was added to the triphenyl phosphite.

The tube was sealed, brought to room temperature and the

contents thoroughly mixed. It was then heated in a thermo-

statted oil bath at 100 °C for 50 h, after which time the mixture became white and crystalline.

The tube was then gradually cooled to -80 °C and opened,

and the volatile materials (unreacted methyl bromide,

2.61 ppm in CDCl^) were removed under water pump pressure

and collected in a cold trap. The crude solid product was

transferred into a sintered-glass funnel with a cap, washed

several times with dry ether, and dried under high vacuum to

yield methyltriphenoxyphosphonium bromide (25.7 g» 90%), m.p. 155-158 °C (sealed tube) (Found: Br, 19.6. Calc, for

C^gH^gBrO^P; Br, 19.7%), (CDCl^) 5.24 (Me, d, I7.4

Hz), 7*55 (phenoxy protons, s), (CDCl^) 41.8 (quartet),

cTq (CDCl^) 10.1 (CH^, d, Jp(. 127.0 Hz), 120.4 (d, J 4.3 Hz),

128.4 (d, J 1.2 Hz), 13 1.3 (d, J 1.2 Hz), 148.8 (d, JpQQ

10.4 Hz) (Ar).

91 1 5 Preparation of Methylbriphenoxyphosphonium Iodide

Redistilled triphenyl phosphite (16,7 g, 1 mol. equiv,) and methyl iodide (11.3 g, 1.48 mol. equiv.) were placed in a round-bottomed flask fitted with a double-surface reflux condenser and a calcium chloride tube. The mixture was refluxed gently at 86-90 °C on an oil bath for a total of

16 h, after which the mixture became viscous and reddish- brown in colour. After cooling for one hour the mixture started to solidify, and in two hours time the crude product

(25*3 g, 104%) m.p. 70 (sealed tube) was obtained. It was then ether washed and vacuum dried to give the purified product as white needle-like crystals, m.p. 115 ^0 (sealed tube), (CDCl^) 3.12 (Me, d, JpQjj 16.8 Hz), 7.44 (phenoxy •j protons, s). The only difference between the H n.m.r, spectra of the crude and purified products was that the former had a peak at cT2.12 assumed to be due to unreacted

Mel, whereas in the latter this was missing. (Before hea­ ting the peak due to Mel in the reaction mixture lies at cT 1.84).

Recrystalliaation of the crude product (3.8 g) by 15 dissolving in acetone and adding ether as described gave creamy crystals of methyltriphenoxyphosphonium iodide

(2.2 g, 57.9%), m.p. 115 °C (sealed tube) (Found: I, 28.0.

Calc, for C^gH^glO^P: I, 28.1%), (Tp (CDGl^) 40.9 (quartet),

(Tg (CDCl^) 10.9 (CH^, d. Jpg 127.5 Hz), 120.2 (d, J 4.1 Hz),

128.1 (d, J 1.4 Hz), 131.2 (s), 148.5 (d, JpgglO.2 Hz) (Ar).

It was later discovered that dry chloroform was prefe­ rable to acetone for use in recrystallisation. Also sunlight

92 was best excluded as this caused the crystals to decompose very quickly and to turn red in colour.

Preparation of Diphenyl Methanephosphonate^^

Crude methyltriphenoxyphosphonium iodide (21.5 g) was treated with 2 M KOH (20 ml) with shaking and cooling. When all the solid had dissolved, the brownish colour disappeared and the mixture was then transferred to a separating funnel.

The oily product was washed further with 2 M KOH (20 ml) and with H^O, and dried over sodium sulphate to yield the crude phosphonate (8.8 g, 77*5%) as an oily liquid.

The product was distilled to give diphenyl methanephos- phonate (6.0 g, 52.8;?^), b.p. 146-I48 °C at 0.3 mmHg, n20 D 1 .5520, rn.p. 35*0 C (sealed tube) (after solidification on standing) (lit.^ b.p. 190-195 °C at 11 mmHg, m.p. 36-37 °C),

(CDCl^) 1.65 (Me, d, JpQjj 17*7 Hz), 7*17 (phenoxy pro­ tons, s), c5^ (CDCl^) 23.8 (quartet), (CDCl^) 11.4 (CH^, d, JpQ 144.6 Hz), 120.6 (d, J 4.4 Hz), 125.3 (d, J 1.1 Hz),

129.9 (d, J 1.1 Hz), 150.4 (d, JpQ^ 8.2 Hz) (Ar).

Attempted Deoxygénâtion of Diphenyl Methanephosphonate by Triphenyl Phosphite

Diphenyl methanephosphonate (O.I6 g, 1.0 mol. equiv.) and triphenyl phosphite (0.20 g, 1.0 mol. equiv.) were sea- 1 1 led in a H n.m.r. tube. The H n.m.r. spectrura of the mixture showed a singlet for phenoxy protons in the aroma­ tic region and a doublet (JpQjj 17.7 Hz) for the methyl protons. The difference in chemical shift of the two was

5.48 ppm. Then the tube was heated in an oil bath, starting

93 at 80 °C, and gradually increasing the temperature in

steps to 150 °C. After 15 h no noticeable reaction had

occurred. Finally the tube was heated continuously in a

thermostatted bath for 305.5 h (175 °C) and 30.5 h (220 °C)

but there was still no change in the H n.m.r. spectrum.

After the addition of TMS the chemical shifts were shown

to be exactly the same as those of the freshly prepared

unheated reaction mixture: cT 1.52 (Me, d, JpQjj 17*7 Hz),

7.09 (phenoxy protons, s). The chemical shift difference

remained the same at 5«45 (CDCl^) 128.1 (s),

23.7 (quartet) and traces at -I7 .I

To make sure that deoxygenation had not occurred in the

above experiment methyl iodide (0.18 g, 2 mol. equiv.) was

added, the tube was resealed and the mixture heated for 16 h

at 100 °C. The n.m.r. spectrum (CDCl^) then showed only

the presence of unreacted phosphonate, <5* 1.69 (Me, d, JpQjj

17.7 Hz) and 7 .18 (phenoxy protons, s), an excess of

unreacted methyl iodide, cT 2.02 (s), and methyltriphenoxy-

phosphonium iodide, S' 3*09 (Me, d, JpQj^ 16.8 Hz) and 7*40

(phenoxy protons, s).

94 Preparation of Tri-o-tolyl Phosphite

o-Cresol (38.3 g, 3 mol. equiv.), petroleum (15 ml, b.p.

30-40 °C) and pyridine (28.0 g, 3 mol. equiv.) were placed in a three-necked flask, fitted with mechanical stirrer, thermometer and dropping funnel. Phosphorus trichloride

(16.2 g, 1 mol. equiv,) in petroleum (25 ml), was added slowly to the reaction mixture with cooling at 0 °C, and stirring. After the phosphorus trichloride was added another portion of petroleum (10 ml) was added and the mixture stirred for 20 minutes at room temperature. Ether

(100 ml) was then added and pyridinium chloride removed by filtration. The filtrate was washed twice with distilled water (50 ml) and the etherate solution was dried with sodium sulphate, concentrated under high vacuum and distilled to give a main fraction (24.3 g, 58.4?^), b.p,

I6O-I62 at 0.01 mmHg, (Found: C, 72.2, H, 6.3, P, 8 ,5 . 20 Calc, for C2^H2^0^P: C, 71.6, H, 6.0, P, 8.8i^), n^^ 1.5780,

6^ (CDCl^) 2.17 (CH^, s), 7.03 (phenoxy protons, m),

J 11.0 Hz), 124.1 (d, J 1.1 Hz), 126.9 (s), 129.9 (d,

J 2.8 Hz), 131.4 (s), 150.4 (d, JpQQ 2.2 Hz) (Ar).

Preparation of Methyltri-o-tolyloxyphosphonium Bromide

Tri-o-tolyl phosphite (15.35 g, 1 mol. equiv.) was placed

in a thick-walled tube and cooled to -80 °C. Methyl bromide

(8.06 g, 1.95 mol. equiv.), also at -80 °C, was slowly added to the ester. Then the tube was sealed and heated in

thermostatted oil bath at 100 °C for 31.75 h after which

95 time the mixture became a white solid. The tube was then cooled to -80 °G, opened, and volatile materials were removed at 15 mmHg.

The product was transferred to a sintered glass funnel, washed with dry ether and dried under high vacuum to give white crystals of methyltri-o-tolyloxyphosphonium bromide

(13.5 g, 69.3%), m.p. 184-187 °C (sealed tube) (Found:

C, 60.0, H, 6.6, Br, I7 .8 , P, 6.4 . C22H2^Br0^P requires:

C, 59.1, H, 5 .4 , Br, 1 7 .9, P, 6.9/0), cTjj (CDCl^) 2.19

JpcH (Tp (CDCl^) 42.1 (quartet), cT^ (CDCl^) 10.9 (CH^P, d,

Jp^ 125.4 Hz), 16.2 s), 119.6 (d, J 2.7 Hz),

128.3-129.3 (m), 133.1 (s), 147.9 (d, Jp^^ 10.2 Hz) (Ar).

Preparation of Methyltri-o-tolyloxyphosphonium Iodide

Tri-o-tolyl phosphite (5.2 g, 1 laol. equiv.) and methyl iodide (3 .17 g, 1.51 mol. equiv.) were sealed in a test tube, which was then heated in a thermostatted oil bath for

19.5 h at 100 °C. The content of the tube became a dark liquid, which solidified after cooling to room temperature. I The tube was cooled, opened and volatile materials were re­ moved under high vacuum. The product was transferred to a

sinterred-glass funnel with a cap, washed with dry ether

and dried under high vacuum to yield yellow crystals of methyltr1-0-tolyloxyphosphonium iodide (5.0 g, 68.5%), m.p. 14O-I48 (sealed tube). Recrystallisation of the product (5 g) by dissolving in

a mixture of acetone and chloroform (1:1) then adding ether

gave white crystals (3.5 g, 50/) m.p. 155-157 ^0 (sealed tube)

96 (Found: C, 52.5, H, 5.0, I, 25.6, P, 5.7. C22H24lO^P requires: C, 53.5, H, 4*9, I, 25.7, P, 6.3%), (GDCl^)

2.20 s), 5.16 (CH^P, d, 15.9 Hz), 7.32

(Ar, s), (Tp (CDCl^) 41.3 (quartet), (CDCl^) 11.9

(CH^P, d, JpQ 127.0 Hz), 16.3 s), 119.5 (d,

J 3.1 Hz), 128.3-129.2 (m), 133.1 (s), 147.7 (d,

10.4 Hz) (Ar).

97 |T% .<« ^ .V»

Thermal Decomposition of Methyltriphenoxyphoaphonium Bromide In the absence of solvent

Methyltrlphenoxyphosphonium bromide (approx. 0.8 g) was -1 sealed in a H n.m.r. tube. The tube was heated in a thermo- statted bath at 175 °C for a total time of 86.5 h, after which the solid became a liquid that even after cooling remained as a clear viscous fluid. The H n.m.r. spectrum showed signals for phenoxy protons centred at cT7 »26, phosphonium protons at (Me, d, 16.8 Hz), 41.6%, unidentified protons (X^) at cT3.20 (doublet, J 15.0 Hz),

28.3%, and phosphonate protons at cT 1 .76 (Me, d, 17.4

Hz), 30.0%, on the basis of signal height. The tube was then heated at 175 for another 4O .5 h and at 200 for

12 h after which the H n.m.r. spectrum showed new signals at cT^2.90 (doublet, J 14.4 Hz), 5.0% for other unidentified protons (X^), phosphonium signals at (T3.70 (Me, d), which had decreased from 41.6% to 9.5%, signals (X^ ) at cf^3.20

(doublet) which had decreased from 28.3% to 28.1% and phosphonate signals at cTl.76 (Me, d) which had increased from 30.0% to 5 7 .3%. The tube was heated again at 200 °C -1 for 52 h. Then the H n.m.r. spectrum showed that the phosphonium protons at ¿^3*70 (Me, d) had completely disappeared, and that protons (X^ ) had decreased from

28.1% to 22.2%, protons (X^) had increased from 5.0% to

9 .0% and that the phosphonate protons had increased from

5 7 .3% to 68.8%. The heating was prolonged at 200 °C (at intervals) for 262 h, and because decomposition was slow, for a further 78 h at 225 °C. Then the H n.m.r. spectrum showed the appearance of more new signals for unidentified

98 protons (Xj), which steadily grew at 1.66 (doublet,

J 14.4 Hz), 1 6 .0%, in the phosphonate area. After further heating for 55 h at 225 °C the n.m.r. spectrum again showed the appearance of new signals (X^) at S' 2.22 (doub­ let, J 15*0 Hz), 1.2%, and the presence of the following signals; (X^), (X2 ), (X^) and phosphonate.

The heating was further prolonged at 225 °C at intervals, for 26 h, and for 32 h at 250 °C, at which stage the 1 I-' H n.m.r. spectrum showed that the signals at o 3.11 (d) due to the protons (X^) had disappeared. The heating was continued for another 30 h at 250 °C, after which the 1 H n.m.r. spectrum showed the following composition based on heights of the signals; cT2.74 (doublet) (X2 ) 9 .7%, cT 2.22 (doublet) (X^) 5.9% and cTl.66 (doublet) (X^) 1 7 .5%.

The phosphonate signals at 1 .76 (doublet) however remained stable at 66.9%. Further heating was discontinued.

n.m.r. (CDCl^) showed S' -I7 .O (s) due to (Ph0)^P=0,

S ' 23.1 (quartet) due to (Ph0 )2P(0 )Me, cT52.9, 71.9, 22.5

(small signal) unidentified.

Identification of the Products of Thermal Decomposi­ tion of Methyltriphenoxyphosphonium Bromide in the absence of solvent

In a number of further experiments the products were 51 identified by a combination of 1 H n.m.r., and P n.m.r. as in the following table.

99 0 t— t— C— 0 0 N • • • • • • • ix) c - c - 0 0 0 0 VO T - m r - m T" m T- m T“ m V“ t o 0) 0 0 c - 0 0 0 a Ì W • • • • • • • ' 0 T— o ■■■11 H PH •d T— 0 0 CD CD 0 0) C\J • • • • • • rH 0 »X) m c - CO t o 0 ü 0 X i O CM CM CM t o H t cd X! Q i E h (I4 'W' 1 W CT» v o CM 0 t o m CD <*H MH • • • • • • •H 0 •H Cd VO 0 T“ m X) er» CM 1 0 w d ^— r - CM t , PQ H t H t H t N • • • £ W H t H t H t + 1 1 1 1 CM Hh T- C7^ 1 X) X) m • • • PQ * 0 CM CM CM o "T t 3 K a ß Ö CM T- X) + 0 B ■ íR 1 1 1 1 • • • 0 0 T- CM CM 0 ft w r * & 1 w 523 bOPH MH CM H t O •H 0 •H cd 1 1 1 1 • H t • H 0 w d CM • H t E l W t o rH CO N VX) 0 0 0 0 0 0 o w • • • • • • • PH LTt m m m m m m S 1 m - r - T“ T - 0 T~ 0 T*“ T~ T— CD O JH , w CT\ C7^ er» er» C7^ CD o PQ • • • • • • • w OsJ CM CM CM CM CM CM CM P-) 0 ■ a 1 'd 0 0 t o m CM t o 0 • »-^ + e • • • • • • • X PH 0 3 v o T- CM CM m X) cr» CÖ ¡s j CM 0 0 CM CM CM CM CM CM E f t ñ 0 ■ p- w a 1 W m c - 0 X) t o -r- Eh K bO m CD 0 t o f e 0 03 d T“ CM T- CM CM t o t o O w

N H- Hd- H" H t Ht CM EH X • • • • • • • < C— c - 0 0 0 t— Ö 1 VO T - CM T - T— T- T-* T- CM T- 0 Pt % w CM CM CM CM CM CM CM 9 PQ Ì-0 • • • • • • • PÍ 0 t o t o t o t o t o _IÛ ___ • a ___ + 1 T i CO 0 t o m (D CD Ht • PH g d • • • • • • • a K> 0 3 c~- 0 CM 0 Ht 0 CD 0 0 X) VO m m Ht t o CM PH 0 ft x¡ j y K Xi 1 03 0 m 0 CD CM >mx* MH • • • • • • • t J 0 •H cd C71 t o t o CD m T“ X) p: 0 03 d CM t o CM CM CM CM cd Pi • 4^ m m X ) CM 0 0 r- 0 CM 0 CM 0 t o T“ PÍ C T3 CM H " CM H " CM CM CM Ht CM CM Ht • ij 0 0 <3j ü C-C— c- c- 0 0 0 0 S• OC- 0 0 0 t— • w ft ü m m m m m m m c— t- 0 Ol 0 c- 0 c— 0 EH m- T~ T- 0 5 <0 CM X) 0 0 Ht CM •íH a 0

100

♦ vo VO t - 1 <0 • • <■*> CD W P , P 1 C— crj 0 r - w ü ' Ö w CO 0 cd ^ 5 r - ^ W T - • KN C- 1 1 ' R •*0 • C\J p a R CM T - • CTN <*- « ^ U N P • ^ X Ph cd Cd • p cd • ß O K N » « * 0 ) I'D 1 1 cd H ’ 0) cd U N ^ rH CM " ^ U N W UN » o (d | •> CNJ s K (D N O ^ ü :s^ KJ t - H- JEJ w O ^ w KN • H - • 0 0 00 v_x 0 VO H- V O H - H - ^ H - H ' • V- • T ~ • T “ • k>o u P i n— y— Q T— ^ÇU P *-3^0^ 00 H- 0 • • • • CT» KN 00 vo T J • ^ (D t - • m T i O CT» P ô ô T T Î vo P VO rH LfN • U • u • u K N CÖ T — 1 K N Cd K N Cd K N cd l ^ • ß N CM t í CM t í CM t í CM bOW I 0) CT CT c T •H S W C M V O C— H- C- ♦ • • 0 N • • • • ^rH mcr» W H- C- V O c - CM t > CM r H C7N "»^i- P t> T- C- T- V O 1- V O T- Ixjcö T - CM • • • • t o ñ P T— T- w ru •>—' 1-3 x-' 0rH Ah ^ 6 p CT» tr~- 0 cd v o • • • • . T i VO CO UN O CT» c- 0 00 • • * W K O f—I •* pH K N CM 0 p UN Cd CO ü T - • . © • ß • W 1 CM ß CM q CM ß b O C M 1 to'" 0 0 ® 0 ^ • H CM C~“ T- t J ‘ ra » C M 1f. M , P»T • p N [>• UN UN TJUD -r- IT» • KN W • • • • y 0) ■ «- i j 0 VO U N H" CM H - CM H - B •H ^ »kT— <5 s LT» 0 0 ^ H - T- 1 0-.-^ 0| + • • • • 1 •H ^ S I p CM CM CM CM ■P • '^1 t3 0 ö T- O p. (D p • • • • H CM ülÖ 0 d PL, O 0 P CM P U N I 0) CM • P • P • H 52; -«d- • 1 1 V O (D V O 0 vo D T - o C - CT» W CT» CQ CT» H • P o « 0 43 43» w CM ö (D k ) 0) N 0 0 0 0) » > ^ 10 ) a w • • • • > T p *0 p i •» + CT» H- 0 0 U N C O U N 0 0 U N i H ' - ' T i* ^ p 00 T- T— T“ T- -r- O T i CM • • • • CO 00 CM KN CM CM C J N p i N 0 43 . {rj . PJ T- O PM £ UN CT» KN O • T - K N • • • • CM T- P K M - - ^ O T“ •H tH ^ VO ^ ;S O P i P o O - ^ • . ü pL, P i P 0 0 U N 0 U N U N O U N O 0 0 0 0 0 B f ^ w h ) O e r ~ 0 CM ^ - O CM U N CT» CM KN H -V O 1 • P (D 0 CM T- CM CM T— CM CM CM r- (M CM CM CM B B y W r ‘ r» H •K) *Ö 'tí ® ß tí d) 0 0 ▼“ 0 0 0 V O V O K N K N V O T- K N B •> •> •> 'O B CM CM 0 CD CM V O 00 y~^ 1 K N K N • T- Cd W W''-' H ^ cd ü o eo E 1

102

t Thermal Decomposition of Methyltriphenoxyphosphonium Iodide in the absence of solvent

Methyltriphenoxyphosphonium iodide (approx. 0.8 g) was

sealed in a H n.m.r. tube. The tube was heated in a ther- mostatted bath at 150 °C for a total time of 71 h, after which the solid became a dark viscous fluid. The H n.m.r.

spectrum (without solvent) showed signals for phenoxy protons

at cT 6.0-8.0 (m), phosphonium protons at 2.5-4.0 (broad

signal), and phosphonate protons at cT 1.2-1.7 (d). The tube was opened, CDCl^ with TMS was added and the tube was resealed.

The n.m.r. spectrum (CDCl^) showed signals at S' 1.80

(Me, d, JpQjj 18.0 Hz) (PhO)2P(0)Me 11.0%, (i^3.25 (Me, d,

JpCH 16.3 Hz) (Ph0)^P‘*’MeI“ 8 7 .5%, new signals at cT* 2.6 (Me,

d, JpQjj 15.0 Hz) due to (PhO)2f ’’Me2 l" 1.5% and cT7.45

(centred) phenoxy protons.

Identification of the Products of Thermal Decomposi­ tion of Methyltriphenoxyphosphonium Iodide in the absence of solvent

In a number of further experiments the products were 1 31 identified by a combination of H n.m.r., and P n.m.r.

as in the following table.

103 N t- c- C-- C- w • • • • • • ♦ • o t~- r- c- c- VO C-- o c- LO O- to f~- 0) C-- T- T- c- ■*- 00 <«- O- ■*- i>- •»- S CO t- • • • • • o u,« T- -t- yr~ 00 VO LO H- to c\J 1 t J • • • • • • H i § ÌR 1 00 to o 00 O LO M w o o T- CM CM to to _ CL, 0) iH ê H o o> 00 CO (O to JO A h H • • • • • cd •H O *H 00 Cd LO 00 LO LO 00 H 0> 00 p] B T- s.^ a m Ph N o VO VO VO VO O 1 * a • • • • • • M ♦ CM m LO LO LO LO -<*• LO to LO CM LO w 1 CM ^ T- T— ^ T- T“ T- 'f— ü s • • • • • • + o CM CM CM CM CM CM 01 Pi 1 n\ 1 TÍ uw o T* B g VO CM CM LO ü PU o 3 ^ 1 • • • • • Pi o o to H" to to to © 1 PU -- M-- ^ Pu to tì H O w t£(«H 1 CO H H- 00 00 00 -P o •H O bOH © • • • • • © "H cd B O T— T~ o a 00 c 01 M Eh N 1 x> M a Pt <0 O O O O H- o Xfl © PU • • • • • • O 1 o > a to LO to LO CM lO VO LO CO C~- LO Pu H CO o © 00 T- 00 -r- 00 -r- 00 -r- T- C-- r- S C\J • rH • • • • • • O - t-- m cd PU p: S o LO LO LO LO LO LO LO bO - W c-1 C- 01

104 cd

m

to

5 t ì 1 w CO 0 bOH • • •H O *H cd cr> 00 0> 00 VO w LT\ N 0 00 w • • LPs ^ T“ KN t o• T- CM• i - CM CM 1 H t o I TÍ 00 t o •> P ^ â t J TÍ TJ -P •i •« ^ 1 w ' to KN K" W)H • 1 W w W 0 *H cd T- OQ CO 0 0 Ü tS '— ' 00 CM to m 1 T“ CM -p T-• CM• • • • I>- T~ C\J o p T~

¿ 0 0 0 0 to*^' to E 0 r~^ er» CM t o ^ S rH o CM T- CM CM w O E-®i P ^ P P O O (T» O • > « ✓

105 // w o bO ß ß m 0) w co •H CO ---- - ■P P vo - p w • 0 P i n LO 00 cr> Cd id O Cd VÛ C\J s • • • • •«J 0 O b ) M 3 CM CM TJ 7 d t— K • P CO ^ 3 •r* CM CO CO 0) • M 3 00 w l O o | •H T- LO LO C7" • M 3 C O LO o k> 0 Qi ß O • • t r co C O CM 3 ^ P B Ü • C3 0 ß «wT ¿ 0 CM 0 1 CO f t CM 0 ß CO m 0 W CO ß s ß (NJ (T\ c n 0 CM p •H 0 • v o 0 • • ü tO • ^ o CO • (H CO l O 0 1 CM M « + LO o CM ß r H O 0 M3 f t 0 f t yr~ P O P S i O o (H CO O p CO 0 N W f t 0 o ß m O M3 0 H" 0 1 O w s i 00 O l CO 0 VÛ l O p • f t ß C3 T“ p o V“ M3 VÛ 0) C3 (Ì4 0 • • • • • • ;s c o w C\J 0 bC l O CM 00 CQ CO 0 4-3 w 0 C3 CO p w O i H « 0 o o p • • B B •H O o 0 N_^ r H o EH • 0 b t> - 0 0 Cd CO s 0 o ß S i O •H T“ l O -P w '—' B W

106 M

W m o w

H K

H

EH

KN O (H o W P o o apiq CO • P K <î

Ä Eh P ^

107 Isolation of Dimethyldiphenoxyphosphonium Bromide

Methyltriphenoxyphosphonium bromide crystals (4.3 g,

10.6 mmol) were sealed in a glass tube and heated for 7 days at 175 in a thermostatted oil bath. After heating the tube the contents became liquid, and on standing for several days started again to crystallise. After 8 weeks of

standing the tube was cooled to -80 °C, volatile materials were removed at 0.5 mmHg, collected in a cold trap at -80 °C, and shown to be mainly bromobenzene, (multiplet) 7.1 to

7.6 ppm. The solid residue was washed with ether and vacuum dried to give a mixture (1.9 g) of (PhO)^P'’’MeBr" and

(PhO)2PM62Br*, The filtrate, after concentration, showed the presence of phosphonate. Recrystallisation of the solid mixture (1.9 g) was carried out by dissolving it in a mixture of 20% acetone (ethanol free) and 80% chloroform (by volume) and adding dry ether to give a product (1.0 g), m. p. (sea­ led capillary) 145 °C for fast heating, and 145-180 °C for slow heating, which was mainly (PhO)2P’’’Me2Br", (CDCl^)

2.95 (Me, d, JpQjj 15.0 Hz), 7*45 (Ar, s). Small signals assigned to the impurities (Ph0)-P'*’MeBr” and (PhO)P'*’Me_Br" were also present. The solid crystals were washed with cold acetone and then with hot acetone (40-45 °C) and were recrystallised from chloroform/ether, washed with dry ether, and dried under high vacuum to give pure dimethyldiphenoxy­ phosphonium bromide (0.6 g, 1.83 mmol, 34%), m. p. (sealed capillary) 200-210 °C, (Found: C, 50.3, H, 4.9. Calc, for

C^^H^gBrOjP: C, 51.4, H, 4.99<), «Tg (CDCl^) 2.95 (Me, d, 6H,

108 «,S 50B L _jf

JpCH H . 7 Hz ), 7.41 (Ar, s, 10H), Îp (CDCl^) 96.6 (septet),

(CDCl^) 12.4 (CHj, d, JpQ 87.4 Hz), 120.9 (d, 4.4 Hz),

127.9 (d, J 1,6 Hz), 131.2 (d, J 1.1 Hz), 149.5 (d,

Jpoc -0 Hz) (Ar).

Hydrolysis of Dimethyldiphenoxyphosphonium Bromide

The nmr tube containing the above product was opened to the atmosphere and hydrolysis slowly took place (50% in

20 days) to give phenyl dimethylphosphinate Ph0P(0 )Me2 >

(CDCl^) 1.63 (Me, d, J^^jj 14.6 Hz), 7-19 (Ar, m),

II5 .6-I3O .2 (m) (Ar). A singlet at 8.0 ppra in the ^H n.ra.r spectrum due to OH of the phenol formed was also observed.

Pure crystals of the phosphonium bromide left in a sample -) bottle over night, became liquid. The H n.m.r. spectrum in deuterochloroform then showed the same signals for

Ph0P(0 )M6 2.

Thermal Disproportionation of Dimethyldiphenoxy­ phosphonium Bromide

Dimethyldiphenoxyphosphonium bromide which was isolated end purified as above was sealed (in small amount) in two 1 H n.m.r. tubes which were placed in thermostatted oil bath. The solid did not melt at temperatures up to 200 °C but began to melt slowly at 210 and the heating was then continued for 7.5 days (190 hours). After this time one of the tubes was cooled and opened in a nitrogen atmosphere.

109

7-lr and then deuterochloroform was added to dissolve the substance (approx, 12- 15% concentration) and the tube was resealed. The H n.ra.r. spectrum showed signals at o 2.89

(Me, d, JpQjj 15.0 Hz) due to (Ph0)2P'*^Me2Br“ (36.5 mole %), at (T 2.59 (Me, d, JpQ^ H . 4 Hz) due to PhOP'*’Me^Br”

(35.3 mole %), and at cT 1,75 (Me, d, JpQjj 17.8 Hz) due to

(Ph0 )2 P(0 )Me (28.2 mole %). In the aromatic region were three multiplets centred at 7.23, 7.35, 7.44 Ppm, and a 31 small singlet at 8.8 ppm. The P n.m.r. spectrum confirmed the assignments for (Ph0 )2P^Me2Br” , cT96,0 (septet), for

PhOP'^Me^Br” , 102.2 (decet), and for (Ph0 )2P(0 )Me, S' 23.9

(quartet), and also showed the presence of (PhO)^P(O), cT- 17.7 (singlet). A small signal at cT 125.1 was probably due to (PhO)^P.

The second tube was heated at 210 for a total of 10-12 days after which time some drops of liquid condensed on the walls of the tube. The drops were separated from the solid mass by turning the tube upside down and the tube was left in a fridge for several weeks. The liquid product and the solid product were separated by cutting the tube in two and were separately dissolved in deuterochloroform in a N2 atmos­ phere. The n.m.r. spectrum of the liquid product showed it to consist of diphenyl methylphosphonate cT 1.77 (Me, d, j 17.7 Hz) (86,8 mole %) and of phenyl dimethylphosphi- nate cTl.70 (Me, d, Jp^jj 14.4 Hz) (13.2 mole %); aromatic protons appeared as a wide multiplet centred at 7.24 ppm.

The solid product contained dimethyldiphenoxyphosphonium bromide S' 2.92 (Me, d, JpCE (28.1 mole %), tri- methylphenoxyphosphonium bromide cT2.64 (Me, d, JpQj^ 14.0 Hz)

110 and then deuterochloroform was added to dissolve the substance (approx. 12- 15% concentration) and the tube was resealed. The n.m.r. spectrum showed signals at S' 2.89

(Me, d, JpQjj 15.0 Hz) due to (PhO)2P'*’Me2Br“ (36.5 mole %), at (T 2.59 (Me, d, JpQjj 14*4 Hz) due to PhOP'^’Me^Br“

(35 »3 mole %), and at cTl.75 (Me, d, Jp^jj 17*8 Hz) due to

(PhO)2P(0)Me (28.2 mole %). In the aromatic region were three multiplets centred at 7*23, 7»35, 7*44 Ppm» and a 31 small singlet at 8.8 ppm. The P n.m.r. spectrum confirmed the assignments for (PhO)2P^Me2Br”, cT96.0 (septet), for

PhOP'‘'Me^Br“, cT 102.2 (decet), and for (PhO)2P(0)Me, cT 23.9

(quartet), and also showed the presence of (PhO)^P(O),

(T-17.7 (singlet). A small signal at cT" 125-1 was probably due to (PhO)^P.

The second tube was heated at 210 °C for a total of 10-12 days after which time some drops of liquid condensed on the walls of the tube. The drops were separated from the solid mass by turning the tube upside down and the tube was left in a fridge for several weeks. The liquid product and the solid product were separated by cutting the tube in two and were separately dissolved in deuterochloroform in a N2 atmos­ phere. The ^H n.m.r. spectrum of the liquid product showed it to consist of diphenyl methylphosphonate cT 1.77 (Me, d,

17*7 Hz) (86.8 mole %) and of phenyl dimethylphosphi- nate cT I .70 (Me, d, JpQjj 14-4 Hz) (13-2 mole %); aromatic protons appeared as a wide multiplet centred at 7*24 ppm.

The solid product contained dimethyldiphenoxyphosphonium bromide cT 2.92 (Me, d, JpQjj 14-7 Hz) (28.1 mole %), tri- methylphenoxyphosphonium bromide cT 2.64 (Me, d, JpQjj 14-0 Hz)

110 b : « j- iìm U sbc. u i .-ti'

(54*2 mole %), and diphenyl methylphosphonate cT 1.77 (Me, d, JpQjj 17*7 Hz) (12.5 mole %), A signal at c^2.19 (Me, d,

JpCH 14.0 Hz) was assigned to trimethylphosphine oxide

(5.5 mole 9é), Aromatic protons appeared as multiplets centred at 7*25, 7 «35 and 7*45 ppm.

Isolation of Trimethylphenoxyphosphonium Bromide containing some Dimethyldiphenoxyphosphonium Bromide -I The contents of the above H n.m.r. tube (solid product) were treated with dry ether. The liquid became cloudy and was placed in fridge over night when a very fine powder was precipitated. Then the liquid was decanted and the solid powder was washed with dry ether in a nitrogen atmosphere and dried under high vacuum to give a mixture of trimethyl­ phenoxyphosphonium bromide (5^ (CDCl^) 2.66 (Me, d, JpQ^

14.0 Hz) (64.0 mole %) and dimethyldiphenoxyphosphonium bromide cT 2.95 (Me, d, JpQj^ 15.0 Hz) (36.0 mole %), Aromatic protons appeared as a multiplet centred at 7*48 ppm.

Hydrolysis of Trimethylphenoxyphosphonium Bromide and Dimethyldiphenoxyphosphonium Bromide

An nmr tube containing (Ph0)2fMe2Br (36.5 mole %),

PhOPMe^Br (35»3 mole %), (Ph0)2P(0)Me (28.2 mole %) and some triphenyl phosphate, in deuterochloroform was opened to the atmosphere for 4 weeks, after which it contained Me^P(O),

1.81 (Me, d, Jpp.„ Hz), cTp (multiplet). POH 14.1 76.8

Ph0P(0)Mep, 1.62 (Me, d, Jpp,„POH 14.4 Hz), 58.0 (septet).

111 . :v - Î.

(PhO)2P(0)Me,

(quartet), and (PhO)jP(O), iS^ -17>7 (small singlet).

Aromatic multiplet signals centred at 7*17 ppni, and a tall

singlet at 8.63 ppin» were also observed.

In a similar experiment a mixture of PhOPMe^Br

(64.0 mole %) and (PhO)oPMe-Br (56.0 mole %) in CDCl., was Z d 3 allowed to undergo slow hydrolysis in a loosely sealed nmr

tube. The composition at various time intervals was

measured on the basis of the methyl doublets (recorded

above) as follows:

Table V

+ ” + (Ph0)2PMe2Br PhOPMe^Br“ Me,P(0) Ph0P(0)Me2 3 (Tjj 2.84 mole cT p 2.5S mole cC 2.10 mole mole n 2.04 Days Height % Height % Height % Height %

- 81 36.4 212 63.6 - - -

6 16 22.4 33 30.8 50 46.7 traces - a 8 16 21 .9 20 18.4 65 59.6 traces

12 8 17.4 6 8.7 42 6 0 .9 6 13.0

18 11 15.8 traces - 103 63.8 22 20.4

b 23 * - - - - 59 64.2 22 35.8 a Small signal singlet at 8.10 ppm. * The tube was fully opened to the atmosphere, b Tall signal (s) at 5.25 Ppm, probably due to (OH). Aromatic protons remained multiplet centred at 7*28 ppm,

112 i- A*

Isolation of Mixture of Tetramethylphosphonium Bromide containing Trimethylphosphine Oxide

Dimethyldiphenoxyphosphonium bromide previously isolated, -1 was sealed (in small amount) in a H n.m.r. tube, which was heated in oil bath for 3 h at 190-220 °C, 25.5 h to 240 °C,

14.5 h to 260 and about 3 h at 290 °C. After cooling and opening the tube, CDGl^ was added and the tube was resealed.

The n.m.r. spectrum showed multiplet centred cT7.27 (tall signal) (Ar), and presence of trimethylphenoxyphosphonium bromide, cT 2.53 (Me, d, JpQ^ 14.0 Hz) (24.0 mole %), cTp 101.9 (m), 13.3 (GH^, d, Jp^ 64.6 Hz), of diphenyl methanephosphonate, 1.74 (Me, d, JpQjj 17*7 Hz) (48.5 mole %),

(5^ 23.9 (quartet), o^^ 11.5 (CH^, d, Jp^ 144.7 Hz), of phenyl dimethylphosphinate, I .78 (Me, d, JpQ^ 14.4 Hz) (14.9 mole %), (T^ 60.3 (m), (not shown), and a further product tetramethylphosphonium bromide (GH^ )^P‘'’Br ” , 2.12 (Me, d,

JpGH 14.7 Hz) (12.5 mole %), d^ 22.6 (small signal), 10.6

(GH^, d, JpQ 55.5 Hz); also d^ -17*7 due to (PhO)^P(O).

Then the tube was opened to the atmosphere, for 8 days, after which time the hydrolysis was nearly complete. The

^H n.m.r. (GDGl^) spectrum showed multiplet centred S' 7.25

(tall) (Ar) and presence of (GH^)^P'‘’Br" (not hydrolysed), cT 1.99 (Me, d, JpQjj 14.7 Hz) (9.5 mole %), cTp 23.4 (small signal), 10.4 (GH^, d, Jp^ 55.5 Hz), of trimethylphosphine oxide (GH^)^P(O), djj '^•9 (Me, d, JpQjj 13.5 Hz) (20.8 mole %), d^ 70.5 (m), cTq 15 .3 (GH^, d, JpQ 68.4 Hz), of Ph0P(0)Me2, d^ (Me, d, Hz) ( mole %), cTp (m). 1.63 'PGH 14.4 15.9 56.5 (GH^, d, Jp(. Hz), and (Ph0)^P(0)Me, 1.75 15.4 95.2 H

113 (Me, d, JpQjj 17»4 Hz) (53.8 mole %), 24.2 (quartet),

(Tq 11.4 (OH^; d, JpQ 144.7 Hz); also 6'^ -17.7 (s) due to (PhO)^P(O).

Then the solvent (CDCl^) was removed at 12 mmHg, and the contents were extracted with D^O, which was separated with 1 1 a pipette into a H n.m.r. tube, for which the H n.m.r. spectrum (no reference) showed Me (doublet) for (Me )^P'‘’Br ” , and Me (doublet) for (CH^)^P(O).

The residue (CDCl^) showed only I .76 (Me, d, JpQjj

17.7 Hz) (90.3 mole %) due to (PhO)2P(0)Me and 1.59 (Me, d, JpQjj 14.4 Hz) (9.7 mole %) due to Ph0P(0)Me2, also

7.22 ppm centred tall signal (Ar). Then the D^O solution was concentrated under high vacuum and DMSO was added for which the ^H n.m.r. spectrum showed (CH^ )^P''’Br “ at (T 1.87

(CH^, d, JpQjj 15.0 Hz) (31.4 mole %), d"p 25.1 (small signal)

(lit."^^ (^p 2 5 .2 , solvent DMSO); o^ (DMSO) 8.93 (CH^, d,

Jp^ 55.5 Hz), and (CH^)^P(O) at S' 1.55 (Me, d, Jp^j^ 13.2 Hz)

(68.6 mole %) (lit."^^ S'^ 1.39, Jg_p 13.2), 5'^ 53.7 (small signal), 5 1.1 due to Ph0P(0)MOp (impurities), (Tp (DMSO)

16.3 (CH^, d, JpQ 68.4 Hz).

Isolation of Trimethylphosphine Oxide

Methyltriphenoxyphosphonium bromide (5.0 g) was sealed in a test tube, then the tube was heated in an oil bath for a total time of 38.5 h at 190 to 260 °C. After cooling it became a mixture of solid and liquid. Then the tube was opened, and after removing volatile materials, small portion of ether was added and the crystals were filtered and washed

114 with ether giving product of (PhO)2P^Me2Br” . (epprox. 0.2 g, weight was not recorded),m.p. (sealed tube) softened at

160 °C, and melted at 180-195 °C. The n.m.r. (CDC1_) j spectrum confirmed the product at 6^ 2.96 (Me, d, JpQj^ 14.7

Hz), 7.47 (s) (Ar).

The ethereal extract after concentration and adding ether gave a second product of grey solid approx. (0.4 g),

(m. p. not recorded, very sensitive to moisture). The pro- -1 duct was dissolved in CDCl^ and was sealed in a H n.m.r. A tube. The H n.m.r. spectrum showed (S' 7*29, 1 t 7*47 (Ar) and a mixture of three products: cT 2.88 (Me, d, JpQjj

15.0 Hz) (23.6 mole %), (Tp 96.5 (m), 12.2 (CH^, d, Jp^,

87.3 Hz) due to (Ph0)2P'*’Me2Br“ , 5*^ 2.60 (Me, d, JpQjj 14.0 Hz)

(52.3 mole %), cTp 102.1 (m), 13.5 (CH^, d, Jp^ 64.7 Hz) due to PhOP'^’Me^Br” , and ¿ ^ 2 . 1 9 (Me, d, JpQjj I .O Hz) H 4 (24.0 mole %), 14.9 (CH^, d, Jp^ 67.7 Hz) due to Me^P(O) the hydrolysis product. Traces of (Ph0)2f(0)Me, (Sp 24.2, and (probably) Me^P'*’Br” , 22.9 were also detected.

The tube was opened and the contents were hydrolysed with a few drops of H2O (in dilute CDCl^ hydrolysis by

atmosphere was complete after 14 days).The H n.m.r. spect­ rum (CDCl^) then showed cT8.56 (s) tall signal, and multi- plet at 6.5-7.8 ppm (Ar), (f* 3.4 (s) for H2O, Me^P(O) at

(T 1.92 (Me, d, JpQjj 14.0 Hz) (74.1 mole %), 74.4 (m),

cT^ (CH^, d, JpQ Hz), and Ph0P(0)Me^ at (T„ 1.69 15.0 68.4 H (Me, d, JpQjj 14.0 Hz) (25.9 mole %), cTp 58.7 (m), cT^ 15.2

(CH^, d, JpQ 94.6 Hz), and small signals for (CH^)^P‘^Br",

cTjj 2.22 (CH^, d, J p Q j j 13.8 Hz) (3.1 mole %) , cTp 21.9.

The solvent (CDCl^) was then removed at 12 mmHg and the

115 residue was extracted with D^O, which was separated with a pipette to give a solution containing trimethylphosphine oxide Me^PiO), (D2O) 1.57 (Me, d, 13.2 Hz) (99°/o),

(Tq (D2O) 18.7 (CH^, d, Jp^ 70.2 Hz), 128.8 Hz) coupling constant in D2O, (lit."^^ C^^-H coupling constant 129 in D2O), ci^(D20) 53.6 (at pH 4), 53.0 (at pH 7) (decet), m.p. 130 °C (sealed tube), 92 (M^, 100), 77 (Me2P0'‘',

100). Isolation of Dimethyldiphenoxyphosphonium Iodide

Methyltriphenoxyphosphonium iodide crystals (I6 .O g,

35.4 mmol) were sealed in a glass tube and then heated for

120 h at175 °C in a thermostatted bath. After cooling the tube, the contents became a dark viscous fluid. Then the tube was opened, volatile materials were renjoved at

12 rnmHg, by collecting in a cold trap at -80 °C, and shown to be iodobenzene (CDCl^) 6 .7-7.8 PPi^ (1^) (Ar), and

2.12 ppm due to (unreacted) methyl iodide. The residue showed signals at (CDCl^) 1.73 (Me, d, JpQj^ 17.7 Hz) due to (Ph0 )2P(0 )Me (47.8 mole Yo), 2.12 (Me, dd) due to

CH^P'*’-0-P intermediate (braces), 2.64 (Me, d, JpQjj 14.4 Hz) due to (Ph0 )2P‘^Me2 l" (18.4 mole %), and 2.83 (Me, d, Jp^^^

16.8 Hz) due to (Ph0)^P'*’MeI'’ (33.8 mole %),

Then dry ether was added to the residue, which precipi­ tated dark crystals a mixture of (Ph0 )2P^Me2 l“ and

(Ph0),P‘‘’MeI” (approx. 6 g). The filtrate after concentra- 3 tion, showed the presence of phosphonate. The solid mixture was recrystallised from acetone/ether several times, and finally from acetonitrile/ether, washed with dry ether, and

116 dried under high vacuum to give pure white needles of

dimethyldiphenoxyphosphonium iodide (1.0 g, 2.67 mmol,

15»2%), m. p. I52-I55 °C (sealed tube) (Found; C, 43.3,

H, 4 .3 , I, 3 3 .9 . requires: C, 44.9, H, 4.3,

I, 3 3 .9%), djj (CDCl^) 2.91 (Me, d, 14.4 Hz), 7.45

(Ar, s), 5^p (CDCl^) 94.8 (septet), 13.2 (CH^, d, Jp^

87.9 Hz), 120.5 (d, J 4.3 Hz), 121.2 (s), 128.0 (s),

1 3 1 .3 (s), 149.1 (d, JpQ^ 11.0 Hz) (Ar).

Hydrolysis of Dimethyldiphenoxyphosphonium Iodide

The H n.m.r. tube containing the above product was opened to the atmosphere and hydrolysis slowly took place

(in 11 days) to give phenyl dimethylphosphinate Ph0P(0 )Me2 ,

1.78 (Me, d, JpQjj 14.2 Hz), 6 .7-7.5 (m), and 8.22 (Ar, s ),

5"p 65.4 (ra), 5'q 15.2 (CHj, d, 95.6 Hz), 115.6-150.5 (m),

149.7 (d, JpQQ 9.5 Hz) (Ar). Pure crystals of dimethyldiphe­ noxyphosphonium iodide in a desiccator became red in 24 h.

The same crystals left in a sample bottle over 2 days, -| became red liquid; the H n.m.r. spectrum (CDCl^) then showed the same signals for Ph0P(0 )Me2 .

Thermal Decomposition of Dimethyldiphenoxyphosphonium Iodide

Dimethyldiphenoxyphosphonium iodide isolated and purified was sealed (in small amount) in a H n.m.r. tube, which then was heated in an oil bath for total time of 28.5 h at

117 190-240 C. After cooling the tube was opened, CDCl^ -I added and the tube was resealed. The H n.m.r. spectrum

showed (T 7»26, 7*40, 7*49 multiplet (Ar), 5 2.80 (Me, d,

JpCH U.4 Hz), 93.7 (ni), 12.9 (CH^, d, Jp^ 87.9 Hz)

due to (PhO)2 P^Me2 l (26.0 mole %), which had not decompo­

sed, 2.60 (Me, d, Jp^jj 13.8 Hz), 5'p 100.4 (m), 14.1

(CH^, d, JpQ 64.5 Hz) due to PhOP'''Me^I“ (36.2 mole %), and

1.75 (Me, d, JpQj^ 17.8 Hz), ¿“p 24.0 (m), 5^ 11.6 (CH^,

d, JpQ 144.0 ) due to (PhO)2P(0)Me (37.8 mole %), Also ¿'p

40.4 (small signal) probably due to (PhO)^P'^Mel“ .

Then the tube was opened to the atmosphere, when hydro­

lysis was very slow (approx. 15-20 days), after which 61 n (CDCl^) showed multiplet lines centred at (S' 7 .15, 7.20 (Ar),

and overlapped signals at S' I .70 (Me, d, JpQjj 13.8 Hz),

5^p 6 5 .7/ (m), (Tq 16.2 (Me, d, Jp^ 69.6 Hz) due to Me^P(O),

1.6 (Me, d, JpQj^ 14.0 Hz), 6^p 59.1 (m), 6'^ 15.7 (Me,

d, Jp^ 92.2 Hz) due to Ph0P(0)MOp, and 6^ 1.75 (Me, d.

Hz), 6*p (m), cTp 11.6 (Me, d, Jp^ 144.0 Hz) PCH 17.8 25.5 due to (Ph0 )2P(0 )Me.

Isolation of Trimethylphenoxyphosphonium Iodide containing traces of Tetramethylphosphonium Iodide

1 A H n.m.r. tube, approximately f filled with methyltri- phenoxyphosphonium iodide, was sealed. The tube was heated

in a thermostatted bath at 210 °G for approx. 120 h. Then

the tube was opened and dry ether was added, which precipi­

tated dark crystals; these were recrystallised from

118 acetone/ether, washed with dry ether, and high vacuum dried to give trimethylphenoxyphosphonium iodide, m.p.

139-142 °C (sealed tube), 6'j^ (CDCl^) 2.65 (Me, d,

13.5 Hz), 7.42 (Ar, s), (Tp (CDCl^) 101.3 (decet), and traces of tetramethylphosphonium iodide, (CDCl^) 2.25

(Me, d, JjPCH Hz), (Tp (CDCl^) 23.3 (small signal)

The Washings, on standing gave a second very small crop of long needles, d'g 1,89 (d, JpQjj 12.2 Hz) (unidentified).

Isolation of Trimethylphosphine Oxide from the Thermal Decomposition of ( P h O ) ^ P'*'MeI~

Methyltriphenoxyphosphonium iodide (7.7 g, 1 7 .O mmol) was sealed in a test tube and then heated in an oil bath for a total of 21.5 h at 190-240 °C. After the tube was opened, and volatile materials removed, ether was added to precipitate dark crystals, which were filtered, washed with ether and vacuum dried to give 1 g of a mixture, m. p. (sea­ led tube) softened 80-85 °C, melted about II5 °C, Si. (CDCl,)

7.43 (m), 7.49 (s) (Ar), 2.76 (Me, d, Jp^j^ 14 .O Hz), cTp 93.2

(septet). S q 13.2 (CH^, d, Jp^ 87.9 Hz) due to (Fh0)2P'^Me2l“

(45.8 mole %), and S'^ 2.54 (Me, d, Jp^^^ 13.5 Hz), cTp 100.1

(decet), 5'q 14.4 (CH^, d, JpQ 64.7 Hz) due to PhOP'^Me^I" (54.2 mole %), The ethereal washings after concentration gave (Ph0 )2P(0 )Me (6 g, 13.3 mmol) .

Then the contents of the H n.m.r. tube were hydrolysed with a few drops of H2O, and the H n.m.r. spectrum (CDCl^) showed (T 6.6 to 7.4 (Ar, m), I.70 (Me, d, JpQjj 13.8 Hz),

<5^p 57.6 (septet), I5.6 (GH^, d, Jp^ 95.8 Hz) due to

119 PhOP(O)M (50.2 mole %), I (Me, d, 13.2 Hz), 05 .79 PCH (Tp 64.6 (decet), 16.4 (CH^, d, Jp^ 69.6 Hz) due to

Me^P(O) (49.8 mole %).

The solvent (CDCl^) was removed at 15 mmHg, and the residue was extracted with H2O (0.5 ml), which was separated and concentrated under high vacuum, to give trimethylphos- phine oxide, (5^ (DMSO-d^) 1,50 (Me, d, JpQjj 13.2 Hz)^, cTp (DMSO-d^) 46.4 (m), cT^ (DMSO-d^) 16.7 (CH^, d, Jp^

68.4 Hz). 42 (a) (lit. 1 .39, Jjj_p 13.2 in DMSO).

1 A similar H n.m.r. tube containing the mixture of

Ph0P(0)Me2 and P(0)Me^ was additionally hydrolysed with drops of KOH solution and extracted with ether. Then the ethereal extract was washed with H^O, dried (Na2S0^), and concentrated to give (CDCl^) 6.5-7.5 (Ar, m) 1.62 (Me, d, Jp^jj 13.8 Hz), cTp (CDCl^) 55.2 due to Ph0P(0)Me2.

The aqueous layer was evaporated to dryness under reduced pressure to give a residue (CDCl^) 1.47 (Me, d, JpQ^

13.5 Hz), (Tp (CDCl^) 40.0 of purified Me^P(O), cTp (CDCl^)

57.3 (at pH 4).

120 Hydrolysis of Methyltriphenoxyphosphonium Bromide in Deuterochloroform

Methyltriphenoxyphosphonium bromide, approx. 4% solution

in deuterochloroform, was transferred to an unsealed 1 -, H n.m.r. tube. Half an hour after preparation, the H n.m.r,

spectrum showed signals at 6" 7.38 (Ar, s), (^3.12 (Me, d, ■ pçjj 17*1 Hz) (phosphoniuia sait, 77 mole %), and (T 1 .75

(Me, d, JpQp 17.7 Hz) (phosphonate, 23 mole %), As time

progressed the phosphonium peaks decreased in size, and

moved from cT 3 .1 2 to S' 2.93, while the phosphonate peaks

at cT 1.75 grew at a constant chemical shift. A second peak

also appeared in the phenoxy protons region at (T7.18 cor­

responding to the phosphonate product and grew steadily,

while the original phenoxy proton signal at cT7.38 (for the

phosphonium salt) decreased in size and slowly shifted to

S' 7.28, After 48 h exposure to air both peaks were equal in height. Finally a peak for the OH group of phenol appeared

at ^ 6.12, and moved slowly to S' 4.85. (Confirmed by adding

a drop of phenol when a single OH peak was obtained at

S' 5 .96). Hydrolysis of the phosphonium salt to the phospho­ nate was 90% complete after 7O h, and 100% complete after

11 days.

In a second experiment methyltriphenoxyphosphonium bromide, approx. 12% solution in deuterochloroform, was 1 1 placed in an unsealed H n.m.r, tube and the H n.m.r. spectrxim was monitored at ambient temperature over a period of 18.5 days. The results were as follows (t/h, phosphonium

121 V

aalt/mole phosphonate/mole 9é): 0, 100, 0; 14, 98.58,

1.42; 22, 97.40, 2.60; 48, 92.31, 7.69; 89, 80.72, 19.28;

1 7 1 , 6 2.29, 37.71; 2 1 8 , 52.22, 47-78; 262.5, 49.40, 50.60;

3 5 7 .5 , 3 7 .4 3. 6 2.5 7 ; 446, 2 7 .5 5, 7 2 .4 5 .

Hydrolysis of Methyltriphenoxyphosphonium Iodide in Deuterochloroform

Methyltriphenoxyphosphonium iodide, approx. 12% solution -1 in CDCl^, was transferred to a loosely capped H n.m.r. tube, djj 3.08 (Me, d, JpQjj 16.5 Hz), 7.43 (Ar, s). After

20 h at room temperature phosphonate signals appeared at cijj 1.80 (Me, d, JpQjj 17.7 Hz) (13 mole %), in addition to the phosphonium signals (8 7.0 mole %) as above. The change of phosphonium compound to phosphonate was observed for 13 days by H n.m.r. spectroscopy. The results were as follows

(t/h, phosphonium salt/mole %, phosphonate/mole %): 20,

86.6, 13.4; 3 9 .5 , 7 3 .3 , 26.7; 64.75, 57.5, 42.5; 90, 48.0,

52; 1 3 7 .5 , 32.2, 6 7.8 ; 162.5, 23.8, 76.2; 186.5 , 18.5, 81.5; 2 1 7 .5 , 12.2, 8 7.8; 240, 10.0, 9O..O; 13 days - 100%.

122 Hydrolysis of Methyltri-o-tolyloxyphosphonium Bromide in Deuterochloroform

Methyltri-o-tolyloxyphosphonium bromide, approx. Q%

solution in deuterochloroform, was transferred to an 1 1 unsealed H n.m.r. tube. After 20 h the H n.m.r. spectrum

showed signals at o 7*26, 7.10 (Ar, m), cT 2.16 (CH^, s),

3.24 (Me, d, JpQj^ 15.6 Hz) (phosphonium salt, 74.2 mole %)

and o 2.24 (CH^, s), 1.81 (Me, d, JpQ^j 17.4 Hz) (phospho­ nate, 25.8 mole %), As time progressed the phosphonium peaks decreased in size, while the phosphonate peaks at

(T 2.24 and S' 1.81 grew at constant chemical shift. The re­ sults were as follows (t/h, phosphonium salt/mole %, phos- phonate/mole %): 20, 74.2, 25.8; 39.5, 69.9, 30.1; 48.33,

6 5.1 , 34.9; 6 3.5 , 5 8 .4 , 4 1 .6; 13 days, -, 100. A peak for the OH group of phenol appeared at ¿"5.27 and moved slowly to ( T 5 . 8 j a peak for CH^ of cresol was recorded at cT 2 . 2 1 (s ).

Hydrolysis in air

Crystals of (CH^C^H^O)^P'*'MeBr“ were placed and stoppe­ red in a sample tube; they became liquid over-night at •| room temperature. The H n.m.r. spectrum (CDCl^) showed that hydrolysis was complete.

Hydrolysis of Methyltri-o-tolyloxyphosphonium Iodide in Deuterochloroform

Methyltri-o-tolyloxyphosphonium iodide, approx. 8 % solu-

123 tion in deuterochloroforra, was transferred to a loosely- -j capped H n.m.r. tube, 3.15 (Me, d, ''^•2 Hz),

2.18 (CH^, s) and 7*31 (Ar, s). After 25 h phosphonate

signals appeared at 2.24 (CH^, s), 1.80 (Me, d,

17*4 Hz) (55.5 mole 9^), in addition to the phosphonium

signals (44*5 mole 9é) as above. The change of phosphonium A compound to phosphonate was observed for 65 h by H n.m.r.

spectroscopy. The results were as follows (t/h, phospho­ nium salt/mole %, phosphonate/mole %): 25.0, 44.5, 55.5;

29.5, 36.4, 6 3.6; 44.5, 3 2 .8, 6 7.2 ; 53.5, 31.5, 68.5;

6 5.0, 28.1, 7 1 .9; 9 days, - , 100. A peak for the OH group of phenol appeared at d" 5*10 and moved slowly to cT 6.20.

124 Thermal Decomposition of Methyltriphenoxyphos- phonium Bromide in Deuterochloroform in a Sealed H n.m.r. Tube

Methyltriphenoxyphosphonium bromide, approx. Q% solution . -1 in deuterochloroform, was prepared in a sealed H n.m.r. tube. The H n.m.r. spectrum, cT 7.39 (phenoxy protons, s),

3.22 (Me, d, JpQjj 17*4 Hz) did not change after 90 h at room temperature, followed by 64 h at 55 °C, 14 h at 95 °C, and 5 1 »5 h at 100 °C in a thermostatted oil bath. The tem­ perature was increased to 125 and new, unexpected peaks appeared centered at cT 2.15 (apparently two closely spaced doublets, each with JpQp 14*4 Hz but subsequently shown to be a doublet of doublets due to the structure P-O-P-Me), about 24.796 in height relative to the phosphonium doublet

(o'" 3 .0 7) after 15.5 h, and 38.896 in height after 60 h. Then the temperature was increased to I50 °C for 37.5 h during which time signals due to diphenyl methanephosphonate,

S 1.75 (Me, d, JpQp 17.7 Hz), appeared.

After a further 46,5 h at 175 °C a new phenoxy signal cT 7,22 (singlet) grew steadily, together with peaks due to dimethyldlphenoxyphosphonium bromide, cT2.74 (Jp^g H.4 Hz).

Heating was continued at I75 °C for another 230 h, when the phosphonium peaks at (T7.33 and at 2.98 (Me doublet) disappeared along with the doublet of doublets at (T2.15.

The phosphonate signals cT 7.21 (singlet), 1.75 (Me doublet,

JpCH 17.7 Hz), and those due to (PhO)2PMe2Br, remained. 51 P n.m.r. (sealed tube) confirmed the presence of

125 (PhO)^P(O) cT-18.0 (16.496), (Ph0)2P(0)Me (T +23-64 (50.596),

(PhO)^P‘‘'MeBr" (5"43.2 (traces), (PhO)2 P^Me2Br” (T95.8

(18.696), and showed peak for unidentified compound at 35»7

(1 4 .596), based on integration of the broad-band decoupled spectrum. -t The course of this experiment was followed by H n.m.r. and the results are shown in the following table.

126 u > (D v o i n i n CT* CM • 0 0 00 r - Xi 1 1 1 1 1 1 1 1 t n i n c - CM T" v o 0 t n t n O to CM t n m i n v o

o . ^ D 0 M w

0 ) IT\ 0 r H f t T" C- CD i n i n v o 0 0 v o e r * 0 0 x> 1 • - p ü ai o CvJ ,ß 1 1 1 1 1 m CD m m T ~ VO 00 CM 0 1 E-* 1 CM m m CM CM m CM CM y— ß f t •H - p 0 ) • s w O O 1 m u i> - PQ • - p 00 0 0 0 m e r * 00 00 T” C\J C\J 1 (D b o 1 1 1 1 1 1 1 1 1 1 0 VO VO t>- v o 0 v o m •H CM CM CM CM CM t n CM LP\ + l o CD ft w -p cd 1 to» ß CM 0 rH PQ CM t n m CM Ov i n 0 m t n CM 0 o (D • - p • 0 p S KN 4:! 0 0 0 0 0 m VO -t— CM T“ 0 i n 0 c - 0 i n 0 ü + b O 0 0 0 0 0 v o v o t > - m CM CM T— ß ft •H T* y— T” 43 ß k> 0 •H m

1 u T-^ ß ÇQ 00 CM CM v o 0 00 r-- v o CM 0) ß • - p S O x i 1 1 1 1 1 1 1 1 1 1 T- v o m CM CM e r * + +» b O i n i n v o t > - 00 00 00 o •H tA u l o 0 ft m O r H > » cr> ß 0 CM t n 00 0 • - p - 4:: 0 0 0 0 0 00 00 m i n i n r - r - 0 O f t b O 0 0 0 0 0 (J^ m CM •H T“ T“ T“ T“ T ~ ß l o 0 O w *H - p •H b O 0 i n i n i n i n i n i n i n co r H ß 0 M ß • • • • • • • O cd * H ß ß ^ ß CD cr> i n v o C7V 00 t > t n i n v o i n i n e r« f t - p - p ß •H 0 v o r CM v o 00 0 CM CT\ T ~ m 0 CM 0 S o 0 0 r H ß T" T“ T- CM CM CM CM m t n i n o EH 0 4:: 0 ü ß 0 0 o 0 CT> ß ß r H to 0 0 cd ß n 4 J ■ P i n i n i n i n i n i n i n i n ß •H 0 ß ft f t • • f • • • • • U - p ß ß 0 0 i n r - t n 00 CTV v o 0 m 00 0 T j - O) cd ■H 0 CO i n 5— CM CM T“ CM CM -T— t r - CM c - 0 + 5 X i W Sm^

1 0 0 U ß ß f t 3 O 0 m i n 0 i n i n i n 0 0 i n i n i n i n i n i n i n ß - P 0 0 0 i n o ^ 0 CM CM CM i n i n r r c ~ - 0 0 0 ß ß T- T- T“ yr" T ~ EH ß

1 - p ü n i n VD t r ^ (T\ 0 T— CM m i n v o i r - 00 0 0 g O T" CM P w ß

127 Thermal Decomposition of Methyl-Triaryloxyphospho- nium Halides in CDCl^ under various conditions

MeP'^(OPh)^Br", MeP"^ (OPh)^l“ , MeP"^ (OC^H^-o )^Br“ , and

MeP"^ (OC^H^-o )^I” were each dissolved in CDCl^ and heated in sealed H n.m.r. tubes. The product compositions were moni- tored by following the H n.m.r. signals for the Me-P protons as described in the previous experiment. Results are shown in the following tables (pp. 128-138). In each case the components are given as mole %,

Thermal Decomposition of Methyltriphenoxyphospho- nlum Bromide (4»59^ in CDCl^ ) at 125

Total heating P'^MeBr" P‘^Me2Br” MeP-O-P P(0)Me time (h)

- 91.5 - - 8.5 8.0 80.4 - 6.7 12.9

24.5 66,8 - 19.4 13.8

29.0 61.2 - 24.4 14.4

32.0 52.8 - 28.2 19.0

48.5 45.7 - 36.9 17.4

68.5 44.2 - 38.5 17.3

9 5.0* 38.5^ - 41 .8^ 1 9 .7^

*) During 8 weeks at room temperature changed to 40.9 44.7 , 14.4 resp., a) 6^p 42.2 (quartet),

b) (JTp 34.04 (d, JpQp 30.0 Hz), 77.40 (dq, JpQp 30.0 Hz, JpQH 14-7 Hz),

c) 6'p 23.89 (quartet).

128 Thermal Decomposition of Methyltriphenoxyphosphonium Bromide (8.0% In CDCI^ ) at 125

Total heating P'^MeBr" P'^Me^Br’ MeP-O-P P(0)Me time (h)

- 96.2 - - 3.8

6,0 90.7 - 4.9 4.4

9.0 88.0 - 6.7 5.2

12.0 83.5 - 8.7 7.8 17.0 81 .2 - 12.1 6.7 32.0 71.0 - 21.8 7.2

41.0 72.2 - 21.0 6.8

48.0 73.0 - 21.3 5.6

61 .0 72.1 - 22.9 5.0

81.0 65.0 - 28.2 6.8

98.0 62.6 - 31.0 6.4

( m in CDCl,) at 125 °C

- 100% - - -

5.0 99.0 - traces -

10.0 94.3 - 5.7 -

12.5 91.0 - 9.0 -

17.5 88.5 - 11.5 -

35.5 83.8 - 16.2 -

38.5 83.0 - 17.0 -

53.5 82.4 - 17.6 -

68.5 83.3 - 16.7 -

112.5* 80.0 - 20.0 -

133.5** 82.7 traces 17.2

*) Heated 44 h at 118 °C **) Heated 21 h at 140 °C

129 Thermal Decomposition of Methyltriphenoxyphosphonium Bromide (4.5% in CDCl^ ) at 175 °C

Total heating P'^MeBr“ P'^'Me^Br” MeP-O-P P(0)Me time (h)

- 92.7 - - 7.3

3.0 51.0 - 36.2 12.8

6.0 47.0 - 38.0 14.9

8.5 47.9 - 3 4 .7 * 17.3

10.5 45.9 - 36.1 18.0 14.0 22.6 2.6 49.2 25.6 16.0 2 1.5 3.3 45.4 29.8 21 .0 33.3 4.0 31.0 3 1.7 24.0 28.9 4.5 36.6 30.0

30.0 37.4 6.5 27.4 28.7 36.0 38.6 7.9 26.0 27.5 50.0 17.6 7.4 37.5 37.5 57.0 27.3 7.9 27.3 37.5 65.0 26.3 10.1 19.7 43.9 77.0 24.7 9.3 24.7 4 1.3

129.0 traces (m) traces tall

201.0 - (m) - tall

302.0 - 9.9 - 90.1

363.0 Including 44 h at 200 ^C; the contents became broi

(CDCl5 ) 23.74 due to (Ph0)2 P(0)Me (89.5%), also (PhO)^P(O) cTp - 17.78 ppm (10 .5%), and (PhO)2PMe2Br

95.75 (traces). *) H n.m.r. after 4 days at room temperature.

130 ThorniQl Docompoaition of Mothyltriphenoxyphospho* nlum Bromide (8,09^ in CDCl^ ) at 175

Total heating P'^MeBr’ P'^MeoBr" MeP-O-P pfol time (h) ^

- 95.8 - 4.2 3.0 47.6 - 45.1 7.2 5.5 44.2 traces 45.2 10.6

(30.0% in CDGl,) at 200 °G j - 100% 0.75 72.0 - 25.2 2.8 1.5 67.3 traces 28.2 4.5 2.5 65.7 1.4 27.2 5.7 3.0 63.3 2.4 27.2 7.1

(a ) and (b) joined together gave

*^P0P ‘^PCH ^ (2 2 .7%), cTp 23.86 (quar­ tet) (2 .09^), 41.75 (quartet) (51.8%), 96.06 (m) (1 .6%) and 33.85 (s) unidentified (2.3%).

(ca. 40.0% in CDCl^ ) at 175

Total heating P‘*’MeBr" P'*’Me«Br” MeP-O-P time (h) 2

6.0 76.6 1.5 19.4

(CDGl^) 41.0 (quartet) due to (PhO)^P'*’MeBr“ (62.9%),

33.5 (d, JpQp 42.65 Hz) (16.8%), 6'^ 76.1 (dq, Jp^p

42.65 Hz, Jp^jj 14.7 Hz) (20.2%), S'q (CDCl^) 8.43 (CH^P, d, JpQ 127.0 Hz) due to (PhO)^P"^MeBr", 13.2 (CH^-POP, dd,

'^PC Hz, ’^pQpQ ^*5 Hz).

131 Thermal Decomposition of Mothyltriphenoxyphosphonium Bromide (10.09^ in CHCl, ) at 125

Total heating P'^MeBr“ P'^Me^Br“ MeP-O-P P(0)M^ time (h) c.

- 1009^ - - -

6.0 97.0 - 3.0 traces

11.5 88.5 - 8.5 3.0

37.5 75.1 - 21.3 3.6

60.5 73.6 - 20.9 5.5

84.5 76.0 - 19.0 4.7

114.0 78.6 - 16.8 4.5

155.0* 79.6 traces 14.5 6.4

*) Heated 41 h at 150 °C m CHCi^, not CDCl used in this experiment.

132 Thermal Decomposition of Methyltriphenoxyphosphonium Bromide (10.09é in CDCl^)

a) in the presence of Triphenyl phosphate (1.2 moL equiv.)

Tempe­ Total rature heating P'^MeBr" P'^Me2Br" MeP-O-P P(0)Me in C time (h)

room - 100% -

125 3.0 100 - «

150 4.5 100 - «

160 8.5 55.2 - 44.8 traces

175 10.5 44.7 - 51.1 4.2 175 15.5 55.2 - 40.5 4.3 175 23.5 44.8 - 46.5 8.7 180 31.0 54.4 3.5 31.6 10.5 170 38.0 33.3 3.0 52.2 11 .5 170 66 • 0 21 .0 8.8 42.1 28.1 170 87.0 23.5 16.2 30.9 29.4 b ) in the nreaence of Triphenyl phosphate (1.5 mol. eouiv. )

room — 100%

160 4.0 57.5 - 36.8 5.7 175 6.0 53.9 - 4 1 .7 4.4 175 11 .0 56.2 - 35.5 8.3 180 19.0 48.1 - 38.9 13.0 160 26.5 50.0 2.5 57.2 10.5 170 33.5 43.2 3.4 37.5 15.9 170 61.5 29.8 6 .4 31.9 31 .9 170 82.5 30.2 13.7 20.5 55.6 c ) in the presence of Triphenyl phosphite (1 mol. equiv.) room 100%

160 4.0 71.0 - 6.9 22.1 180 14.0 66.3 - 10.5 23.2 170 28.5 52.8 1.9 13.9 31.4 170 56.5 50.9 1.8 10.5 36.8

133 Thermal Decomposition of Methyltriphenoxyphosphonium Iodide (8.09^ in CDCl^)

at 125 C

Total heating P'^MeBr” P'^Me^Br“ MeP-0-P P(0)Me time (h)

room 88.7 - 11.3 * room - 79.6 - 20.4

12.0 70.7 - 9.1 20.2

21.0 64.5 - 16.6 18.9

33.0 62.3 - 20.9 16.8

42.0 47.0 - 22.5 30.5

64.0 49.5 - 19.1 31 .4

76.0 57.8 - 18.4 23.8

103.0 39.9 — 20.8 39.3

*) On standing for one year in sealed tube

at 125 C

room 100% - -

15.0 81.5 - 18.5

22.5 82.7 - 17.3

38.0 77.8 traces 22.2

62.5 79.3 traces 17.5 3.2

at 175 °C

room 100% - -

0.5 72.2 1.7 22.7 3.4

134 0)|4c s k TÍ-

O o CO o o ir> VO VO CO (D • • • • • • • • • 00 r—1 o CM VD m VO m 00 O T~ T“ T“ CM CM CM CM t A ß

k>

w

CM

a 0) i n t n 00 c n 00 • • • • • » VO c r \ c ^ t n i n t n T— T- CM CM CM CM

o

I t N ^ » bO 03 rH o o t n i n t n VO i n T" o o | • • CÖ 0) • • • • • • • - p ß H o 00 t n 0 0 C-- C-- 00 0 0 i n i n t n bO •H * B ß 03 ♦ •H ♦ bO r> ß

00

03 rH O O ft O 03 S o ß -p CM o o • ,ß O CM o Nj- in Ü ß -p c- bO B e • • • • • • • R •H o •H in t n t n i n i n M U

bO cd Cd cd cd cd ,Q ft ß <0 o rH *H O o O O o o O o 0) ß Cd

135 O

m m CQ rH 0 C d CO (H ß U bO (U ß •H C\J 03 e n o T- CM o o W s C\J O C d O ß Ü bO •H ß 0 o

ß 0 I 4( N 0 ß u w 0 CQ. ß C d 00 t o o o t o NO bO O s i • X o + l O 0 o o CM o 00 cn l O cr> O p P h T- r H o ON 00 00 NO LO CM 43 •H bO ß a CNJ •rH C d bO rH ß ß C d 0 C O C O X > to o o NO 0 43 H CT» O X CM CM x î •H CO C\J 0 V) ^ ß 0 0 O x î ft

ß Ü t o 00 VD t o CT» CO VÛ ß to o rH O ^ • ü ß o 0 H ß O bO rH ü ß ai cd X r H * H l o etí -P 0) ^ m ir> P. 5 4 3 CÖ g w 0 o O

136 <ü I • CU •6) ^ 0) te H* 1 •H o O) t í CU O n , t í f e : ¿ li 0.'« •H XÌ 0) ITN 0) +> 'V_✓ tí T í X i +a KN tí 0) o O t í +> A j- CU o T í r—t t í 0) cu + J tí --— s tí tí 0) T í O tí B <¡) NW»' «M CU tí O tí k> (0 cu CM • bn S: • CM t í CM •H o CM 4 i t j 03 m Cd cu fH T í ^---V cu t í (0 < t í j t í •H 0) Cd r i • o t í g cg NA « t— í cu CU t— í 1 o x i t í • t í a - p NA pq t í t í cu c o t í t í O • «M o ~ T T +> + NA r—C o • ^ tí Ö NA CU o o • vo o t í O ( t í tí m o cu LT> w Oh • cu hTN (U VO CU tí x i CJ 5 ) -P H-> NA (d o o CU cu o K ^ g l -p rH rH 8 0) O •H cu a tí O tí > - p TÍ NA • A cu ^--- N rH T í tí N CU T í t d CU ä UTA 0) 03 • (t í CÍA o ' P f NA V_✓ cd A i- ( t í co fH •~3 CM •H 10 • CM tk O TÍ A í- t í 04 C\J v o m i r » iT v OJ o ir v ___' tu 3 • • • • • • • • * • w-■ VO O O o t— LP\ 1 ^ cvj C\J iH t — e-f 00 O rH VO v o v o VO VO VO VO ir v • NA O O VO w U í t í r - CJ 1 __^ CU rH i t í y---- s o rH g M N t í n » O cu 0) r-~ co IT\ ir \ líA 00 O r - UA -P UA t í t í • • • • • • • • • • • T í Eh 'S) O v CTN CM <«;}■ ir v VO CO 00 CM CU UA CM CM NA NA NA NA KA NA NA t í NA VO • T í • (0 *~3 fH * o m 03 CsJ T Í T Í

ai NA o r ^ co CM 00 ir \ rH NA l/A VÛ h - OA o rH CNJ • • • • • • • • • CM O O O o o O rH iH rH EH cu CO N A rH fH o S’ CJ a ON (tí O v o C3A «c^- L /> VO t— ON iH K > IfN • • iH a J-

«H ^1

157

‘/v H

O) ' ' ■' ' TJ p H 0 N S a s 5 1 t e >> EH • 9k • P < o P • - «si- O -P f i O O T“ CM t o P P V û T - n T“ •H •H a O D û 0 • .H t o ♦ «d- TJ t e T - 0 tí O , b O P 3 w O B O (U r i O Pi P O t- 0 0 P 0 f i Ph 1-

CM LO TJ g f>3 - T“ «;*• 0 B ü O C\J O A 1 — CM t o •«Í- l O l O as 0 P 1 T- 6 T“ O f i H P i O ^ • 0 O P i P ü lO rO 0 -P C\J O V û f i H 1 N 0 P 0 O H . t e O rO P :i 0 O -P g P T> te s o ^ -P 0 V û H + • (T> f i ü t e P i LT> ! e • 0 Eh 03 KN O t A T - to V û T" CM lO f i t e • • • • • • LO E h gl P i O 0 O O T ~ T— CM LO ■«à- LO TJ C û r H O O V û l O to CM g O t e O O ;Q t e V û c - B P • OJ P i ü T- b to • P J x i t e K N 0 ■P <3 O O , P 'w ' P b O as -P •H B P 0 0 Pi tS 03 O P f i Pi O (T\ CM O C-- • •H rH |ii -P • • • • • • 03 00 03 Cd O O CM P O O CM to LO T— •«d- fH «M- e to bû O 1 O V û LO LO lO lo 0 iH bO I t e «ai • •H P S 0 •H H O 0 b O O P 03 0 H •H P O E h k> 03 S P i Cû ü TJ •H rH KV O 1 0 SP •H P TJ 0 P •H •H •P 0 TJ Pi P ü O -P O O -P O O O O O O 0 0 O e -P 0 t e cd CM t e a s O O O O O O B Td te O 0 cd 0 0 CM CM CM CM CM CM O 00 P i V û E h P P P P • O § O g 0 « d - K> LO 0 tei P 0 lO lO O lO O O O P ü ol S • O P s •H 1 V û 1 O CM «d* V û ■ P 0 E h CM 0 P g b O CM 0 w P • 0 E h 1 0 t — P i 0 +> • P TJ f i LO Q i ü O & «a i 03 0 g t e T“ CM CM to LO V û B P i P O Pi C O U 0 P ü 0 03 «M

158 Hydrolysis of P-Q-P Intermediate obtained from Methyltriphenoxyphosphonium Bromide in Deutero- chloroform

The product obtained by heating a concentrated solution

of (PhO)^P MeBr in CDCl^ for 98 h at 125“140 was divi­

ded and used for two further experiments: (a) for hydroly­

sis by atmosphere, and (b) for attempted isolation of the

P-O-P compound.

(a) The tube was sealed with parafilm and observed by 1 H n.m.r. spectroscopy. After 7 days no change had occurred •j but after 14 days the H n.m.r. spectrum showed the first

stage of hydrolysis (See Table IX a).

(b) The solution was treated with dry ether to precipi­

tate a small amount of white crystals which were removed by 1 filtration under nitrogen. The H n.m.r. spectrum (CDC1-) 5 for the crystals showed the same signals as for the solu­

tion, before separation of the crystals, viz. S 2.16

(MeP-O-P, dd, J PCH 14.5 Hz) (9.5 mole %), 3.1 (Me, d. JPQjj 16.8 Hz) due to (PhO)^P'^MeBr“ (79.0 mole %), and

1.75 (Me, d, 17.7 Hz) due to (Ph0 )2P(0 )Me (11.5 mole %). •j The contents of the H n.m.r. tube were transferred into a small flask, parafilm sealed, and stored in desiccator

(CaCl^) for 13 days after which time the ^H n.m.r. spectrum showed the seme signals as above for the first stage of hydrolysis (See Table IXa). After two more weeks at room temperature the second stage of hydrolysis took place

(See Table IX b).

In a further similar experiment the product from the

139 second stage of hydrolysis in CDCl^ (after 6 weeks) showed •i H n.m.r. signals at 5” 1.86 (d, J 15.0 Hz), unidentified

(X^ ) 2.82 (d, J 18.0 Hz), unidentified (X^) 3.4'Ji,

3.17 (d, J 18.0 Hz), unidentified (X^) 3.4?^, and 1.72

(Me, d, JpQjj 17 .7 Hz) due to (Ph0)2P(0)Me, 82.1^, in the aromatic region multiplet appeared at 6.5 to 7.8 ppm. The 31 P n.m.r. (CDCl^) spectrum showed a strong signal at

6 24.8 (MeP, q, JpQg 17.6 Hz) due to (Ph0)2P(0)Me (63%) and two smaller signals of approx, equal intensity (18-19% each by integration) at 46.7 and 12.0 ppm, appearing as doublets (J £a. 4 Hz) in the broad-band proton-decoupled spectrum and assigned to the P-O-P structure. In the fully coupled spectrum the signal at 46.7 consisted of six lines

(interpreted as two overlapping quartets) (JpQpj 16.0 Hz) and that at 12.0 appeared as a triplet (J 20.3 Hz).

140

f ' '(«»©--y

Thermal Decomposition of Methyltriphenoxyphosphonium Iodide and simultaneous distillation of products

Pure crystals of methyltriphenoxyphosphonium iodide

(13.0 g, 10.6 mmol) were used for simple vacuum distilla­

tion. The apparatus (before use) was flushed with nitrogen.

Then the compound was heated for 6.5 h under 0.2-0.4 mmHg

pressure (bath at 200 °C). The crystals melted to give a

very dark liquid, which after cooling remained dark and

viscous. A cold trap collected iodobenzene (few drops),

(5 6 .7-7.9 (m), an unidentified component, (T 5.2 (s) and a

trace of (probably) Mel,

in ice) also collected a few drops of iodobenzene,

S' 6.2-7 .7 M,

The n.m.r. spectrum (CDCl^) of the residue showed

the presence of phosphonate (18.0 mole %), S I .77 (Me, d,

JpCH 1 7 »7 Hz), and of undecomposed phosphonium salt (82.0

mole %), S 3» 11 (Me, d, JpQjj 16.8 Hz). Signals for phenoxy

protons appeared at (T 7*48 (tall singlet) with a small

multiplet of lines at S 6.98-7.5 ppm. Distillation was continued for a further 6 h under

0.2-0.4 mmHg pressure, at a bath temperature of 230-240 °C.

Two cold traps collected iodobenzene, S 6 .7-8.1 (m), containing a very small amount of impurity, cT 3 .1 2 , whilst

the receiver collected a mixture (approx. 1 g) of liquid

and a small amount of red solid. The liquid contained mainly the phosphonate, (PhO)2 P(0 )Me, (5^ (CDCl^) 1 .67 (Me,

d, JpQjj 17.7 Hz), (CDCl^) 25.0 (quartet). A The H n.m.r. spectrum (CDCl^) of the distillation resi­

due showed signals at S 1.75 (Me, d, 17*7 Hz) due to

142 phosphonate (61,196 in height), S ' 2.12 (MePOP, dd, JpQjj

14*4 Hz) due to the P-O-P intermediate (10.396 in height),

S ' 2.79 (Me, d, JpQjj 14.4 Hz) due to (PhO)2P‘*’Me2l“ (9.5°/6 in height) and cT 3.0 (Me, d, JpQjj 16.8 Hz) due to undecomposed

(PhO)^P'‘‘MeI“’ (19.196 in height). Phenoxy protons appeared at 6.7-7.8 (m). P n.m.r. (CDCl^) confirmed the assign­ ments for (PhO)2 P(0 )Me (38.9 mole 96), cT 23.6 (quartet),

(PhO)^P'*‘MeI” (17.8 mole 96), S' 40.4 (quartet), (PhO)2P^Me2 l~

(5.0 mole 96), S' 93.9 (septet), and for the P-O-P interme­ diate (29.4 mole 96), S' 33.6 (d, JpQp 45.3 Hz), 75*7 (ni,

Jpop 45.3 Hz), and also showed the presence of (Ph0)^P=0

(2.7 mole 96), (T-18.1 (s ), and of (PhO)^P (6,2 mole 96),

S' 127.7 (s).

Then the residue was washed with dry ether. The oily mass did not crystallise, but the concentration of phosphonate was reduced (signals 36.696 in height). Attempted crystalli­ sation from dry chloroform, by adding dry ether gave only an oily mass. The H n.m.r. spectrum showed signals for the phosphonate (13.896 in height), the P-O-P compound (24.696 in height), (PhO)2P^Me2 l” (21.496 in height), and (PhO)^P'*'Mel“ (not decomposed) (4O.I96 in height). No further separation could be achieved.

143 Hydrolysis of the P-0~P Intermediate Formed during Thermal Decomposition of Methyltriphe- noxyphosphonium Iodide in the absence of Solvent

The distillation residue after volatile products had been removed was washed with ether and then recrystallised

(a) from CHCl^, (b) from acetone, the latter causing some hydrolysis to occur. Two stages of hydrolysis were observed for solutions of the distillation residue in CDCl, in -i unsealed n.rm.r. tubes. H n.m.r. data are given in Table X. 31 P n.m.r. data (CDCl^) were as follows:

First stage of hydrolysis

(T 24.1 (quartet) due to (PhO)2P(0 )Me, 40.6 (quartet) due to (PhO)^P‘‘’MeI” , 93.7 (m) due to (PhO)2P'^Me2 l” , 76.0 (d,

Jpop 43.3 Hz) and 33.4 (d, JpQp 40.0 Hz) due to unchanged

POP intermediate (v. small), -17 (s) due to (Ph0),P0,

-25 (s) assigned to (PhO)^P'‘’I“, and unidentified peaks at

4.98 (t, J ca. 20 Hz), 84.4 (m), 96.8 (m), and -11.9 (s).

Second staple of hydrolysis

S' 25.0 (quartet) due to (PhO)2 P(0 )Me, 91.9 (m) due to

(Ph0 )2 P'*^Me2 l" (small), unidentified peaks at 11.7 (triplet,

J 23.3 Hz), and 47.7 (m» possibly two overlapping quartets)

(cf. hydrolysis of corresponding bromide, p.140), and very small signals at -5 (s), and -12 (s).

144 X N O) ^00 m tn r- cd

CD •H N (0 w C\J 00 00 o IT\ t i

0)

0) 0) oco p< o u (U -tí

N—' N 0) ONw 00 o • • CM t ^ a> líN -tí N CO ITNw C J O (U hrN I

01 t)

t— t i N a co ir>Ä 4-» cd • • CM ^

(d I - p t i 01 ' 01 otí » 0) co

145 Thermal Decomposition of Solid Methyltri-o-tolyloxy- phoaphonium Bromide and Distillation under High Vacuum

Pure crystals of methyltri-o-tolyloxyphosphonium bromide

(3.0 g, 0.007 mol) were placed in a small flask protected from moisture with a calcium chloride tube and heated in

an oil bath (200 *^C) for 2 h at 0.1 mmHg. A cold trap

(-80 °C) collected a few drops of liquid which were shown to contain o-bromotoluene {19.6%) (CDCl^) 2.35 (CH^, s) and o-,cresol (20.4%) (S'-^ 2.20 (CH^, s), 4.29 (OH, s).

Aromatic protons were recorded in the region 6.6 - 7.6 ppm (m). During a further 3 h a clear liquid distillate

(1.8 g), b.pt. 160-180 at 0.1-0.05 mmHg was slowly collected in a second cold trap (-80 °C). N.m.r. spectro­ scopy (CDCl^) showed the distillate to consist of di-o-tolyl methanephosphonate (52.4%) c5^ 1 .78 (MeP. d,

JpCH 17.4 Hz), 2.21 (o-GH^, s), cTp 23.85 (quartet), o-bromotoluene (18.2%) 2.34 (CH^, s) and o-cresol

(2 9.4%) 2 ,17 (CH^, s), based on height of the tolyl CH^ signals. Aromatic protons appeared at cT7.10 (m). 31 The P n.m.r. spectrum also showed small unidentified signals, cTp 13.3, 10.9, - 8 .4 , and -11.5 ppm.

G.L.C. analysis of the contents of both cold traps was carried out on a 1 cm x 5 m glass column with 10% PBGA on

Celite (60-80 mesh), at 115 and I7 psi N2 pressure. In each case o-oresol (tj^ 5 niin 35 sec) and o-bromotoluene

(tjj 38 m in 50 sec) were positively identified. No m- or p-bromotoluene (tj^ 42 min 10 sec for both) were detectable.

146 laolation of MeP-0»P Intermediate after Thermal Decomposition of Solid Methyltri-o-tolyloxyphos~ phonium Bromide and Distillation of Volatile Products

The H n.m.r. spectr\im in deuterochloroform of the dark- brovni hard solid residue (0.5 g) from the above experiment showed signals at 6* 2.00 (CH^C^H^, s) (unidentified, 47 »896 in height), 2.15 (GH^C^H^, s) (unidentified, 33.1% in height), 2.24 (CH^C^H^. s) for di-o-tolyl methanephospho- nate (18.9% in height) and 7*17 ppi^i (Ar, s). A small doub­ let centred at cT 2.45 Ppm corresponded to one region of the Me-P-0-P signal (J Hz), the other half being POPCH 1.9 hidden beneath the tolyl phosphonate signal at (T2.24»

After 6 weeks (tube unsealed) the appearance of three small singlets at 4 .88, 5.20, and 5.53 ppm, indicated the products of hydrolysis of a P-O-P complex. Similar signals were found after hydrolysis of the Me-P-O-P intermediate from methyltriphenyloxyphosphonium bromide. See p. I4I.

The total residue was then dissolved in chloroform and treated with dry ether, which precipitated a small amount of white solid and also an oily mass. The solid was decan­ ted with ether, separated, washed with ether, and dried under high vacuum to give white crystals which became dark in colour on standing, (CDCl^) 2.02 (CH^C^H^. d, JpQQQQjj

0.9 Hz, 9 H), 2.18 ( ^ 3 C^H^, d, JpoccCH ^ (Me-P-O-P, 13.9 Hz, JpopcH ^ H), 7-20 (Ar, s, 20 H), cTp (CDCl^) 32.3 (MePOP, d, Jp^p 32.4 Hz), 77*0

(MePOP, dq, Jp^p 32.4 Hz, Jp^^jj I6 .O Hz).

147 III,2.1 OBSERVATION OF THE COURSE OF THE REACTIONS BETWEEN PHENACYL HALIDES AND PHOSPHORUS (III) ESTERS IN DEUTEROCHLOROFORM BY ^H N.M.R. SPECTROSCOPY

(a) Trineopontyl Phosphite and Phenacyl Broraide

Concentrated solutions (ca. 25%) in deuterochloroform of trineopentyl phosphite (0.155 g, 1.0 mol. equiv.) and of phenacyl bromide (0.105 g, 1.0 mol. equiv.) were sepa- -j rately prepared and were mixed together in a H n.m.r. tube. The H n.m.r. spectrum 15 min after preparation showed signals for the following;

Starting materials (57.2 mole % ); phosphite, cT 0.91

(Me^C, s), 3.44 (CH2OP, d, 6.3 Hz); phenacyl bromide,

4.39 (CH^, s).

Arbuzov intermediate (10.6 mole % ), S' 1.02 (Me^C, s).

4.34 (CH^OP"^, d, JpQQjj Hz), (CH^P“^, d, J 4.8 5.74 PCH 18.3 Hz).

Arbuzov product (18.4 mole % ). d 0.86 (Me^C, s), 3.72

(CH^OP, d, JpQQjj 4.8 Hz), 3.65 (CH2P, d, Jp^^^ 22.8 Hz).

Perkow product (13.7 mole % ), ó'0.93 (Me^C, s), 3.78

(CH2OP, d, JpQ^jj 5.1 Hz), 5.3 (CH2=C, m).

Neopont^l bromide (ca. 30 mole % ). cT I .03 (Me^C, s).

148 -1 ■

3.27 (CH2 , 3 ) .

Aromatic protons appeared as a multiplet (7 ,3-8.4 Ppm) 1 The H n.m.r. spectrum was re-examined at intervals and recorded (table No. XI). After 23*5 h only traces of star­ ting materials remained together with the Arbuzov inter­ mediate (4.6 mole %), the Arbuzov product (49.5 mole %) and the Perkow product (46.0 mole %),

149

(b) Dineopentyl Phenylphosphonite ^nd Phenacyl Bromide

Dineopentyl phenylphosphonite (O.I7 g, 1.0 mol. equiv.)

and phenacyl bromide (0.12 g, 1.0 mol. equiv.) were dissol­ ved separately in deuterochloroform (ca. 25-30%) and the

-1 solutions were mixed and transferred into a H n.m.r. tube.

The H n.m.r. spectrum immediately after preparation showed

several signals due to:

Arbuzov intermediate (41 mole % ). cf I.07 (Me^C, s),

4.34 (CH20P'^, m), 5.99 (CH2 P'^, d, 15.6 Hz).

Perkow product (57 mole % ). d 0.89 (Me^C, s), 3.75

(CH2 OP, d,JpocH 5-1 Hz), 5.18 (P0C=CH, d,JpoccH

Neopentyl bromide (ca. 55 mole % ). (product of decompo­

sition), cTl.OO (Me^C, s), 3.22 (CH^, s).

Unreacted starting materials (approx. 2 mole % ). cT 4.45

(CH2 , s) for phenacyl bromide and cT 0.92 (Me^C, s), 3.4

(CH2OP, m) for the ester (traces only).

The Arbuzov product and Perkow intermediate were not detected. No further change was observed after 3 days.

(c) Neopentyl Diphenylphosphinite and Phenacyl Bromide

Neopentyl diphenylphosphinite (0.203 g, 1.0 mol. equiv.) and phenacyl bromide (0,1484 g» 1.0 mol. equiv.) were dissolved separately in deuterochloroform (ca.20-30%) and the solutions were mixed and transferred into a H n.m.r. tube. The H n.m.r. spectrum of the mixture 15 min after preparation showed signals for:

151 Arbuzov intermediate (68 mole % ), 6 0.99 (Me^C, s).

4.15 (POCH,,, d, J 4.2 Hz), 6.25 (CH P, d, 12.6 Hz). POCH 2

Perkow intermediate (10 mole % ),

(CH2OP, d, JpQQjj 5.4 Hz), 5.48 (CH2=C, m, 5.41-5.54).

Perkow product (13 mole % ). & 5.10 (CH2C, m, 5 .0-5.2).

Neopentyl bromide (ca. 13 mole cf 3.21 (CH2 , s).

Unreacted ester (8 mole 9^). <5" 3.4 (CH2OP, m) and phenaoyl bromide, ci 4.8 (CH2 , s).

Aromatic protons gave a multiplet between 7*2-8.35 ppm.

After 3 h the vinyloxyphosphonium (Perkow) intermediate was already decomposed, but the ketophosphonium (Arbuzov) intermediate, which seemed to be stable, did not show decomposition after 28 hours.

(d) Trineopentyl Phosphite and Phenacyl (Jhloride

Concentrated solutions (ca. 25-3050 in deuterochloroform of trineopentyl phosphite (0.155 g, 1.0 mol. equiv.) and of phenacyl chloride (0.082 g, 1.0 mol. equiv.) were sepa- rately prepared and were mixed together in a H n.m.r. tube. A The H n.m.r. spectrum, 10 min after preparation showed signals for the following;

Perkow product (35.3 mole 9^), cT 0.95 (Me^C, s), 3.78

(GH2OP, d, JpocH (CH2=C, m, 5 .14-5 .3 4 ).

Neopentyl chloride (ca. 35 mole 9^), cf O .96 (Me^C, s),

3.29 (CH2 , s).

Starting materials; trineopentyl phosphite (64.6 mole %), cTO.91 (Me^C, s), 3.44 (CH2OP, d, JpQCH phenacyl

152 i >■ , ,

chloride, S 4*68 (CH2 » s).

Aromatic protons appeared as a multiplet between 7.2-

8.10 ppm. The reaction was followed by H n.m.r. (see table) and was shown to be fast; no signals for a phospho­ nium intermediate were observed.

After 2.58 h the reaction was 93.2% complete.

Results of reaction between trineopentyl phosphite and a-chloroacetophenone (1:1 molar ratio) in deutero- chloroform at 36 °C. Product composition

Time (h) (R0)3P (R0)2P(0)0C(

0 .17 64.7 35.3 0.50 35.6 64.4 0.67 3 1.3 68.7

0.83 24.4 75.6

1 .3 3 16.7 83.3 1.50 15.4 84.6 1.7 5 12.2 87.8

2.58 6.8 93.2

24.42 - 100.0

48.00 « 100.0

(e) Dineopentyl Phenylphosphonite and Phenacyl COiloride

Dineopentyl phenylphosphonite (0.30 g, 1.0 mol. equiv.) and phenacyl chloride (0,16 g, 1.0 mol. equiv.) were dis- •j solved in deuterochloroform (ca^. 0.6 ml). The H n.m.r. spectrum immediately after preparation showed that the reaction had already taken place, to give the following:

l-Phenylvinvl neopentyl phenylphosphonate (Perkow pro-

153 duct), cT 0.89 (Me^c, s), 3.77 (CH^OP, d, 5.1 Hz),

5.19 (CH2=C, d, 2.5 Hz), 7.15-8.10 ppm (Ar, m).

Neopentyl chloride. (S' 0.96 (Me^C, s), 3.22 (CH^, s).

A small signal remained for unreacted phenacyl chloride,

(T 4.6 (CH2 , s). The H n.m.r. spectrum was re-examined

after 22 h, but showed no change.

(f) Neopentyl Diphenylphosphinite and Phenacyl Chloride

Phenacyl chloride (0.1231 g, 1 mol. equiv.), dissolved

in deuterochloroform (ca. 0.5 nil), was added in a nitrogen

atmosphere to neopentyl diphenylphosphinite (0.2168 g, 1.0 mol. equiv.)I and the solution was transferred to a 1 1 H n.m.r. tube and sealed. The H n.m.r. spectrum 10 min

after preparation showed signals for:

Starting materials; ester (20 mole 9^), cT 0.93 (Me^C, s).

6.3 Hz), and phenacyl chloride. 3.45 (CH2OP, d, JPOCH (T 4.86 (CH«, s).

Perkow intermediate (35 mole % ), S' 0.98 (Me^C, s), 4 .10

(CH20P'^, d, JpQQjj 4.8 Hz), 5.55 (P'^OC^CHp, m, 5.49-5.61).

Perkow product (vinyl diphenylphosphinate) (45 mole %),

CT5.15 (CH2=C, m, 5.06-5.24).

Neopentyl chloride. cT 0.96 (Me^C, s), 3.26 (CH2 , s).

Aromatic signals appeared between 7.10-8.2 ppm. The

^H n.m.r. spectrum after 20, and 30 min showed decreasing

amounts of the phosphonium intermediate; after 1 hour

traces only remained.

154 III.2.2 PREPARATION AND ISOLATION OP PHOSPHONIUM INTERMEDIATES

The Arbuzov intermediates

Preparation of Trineopentyloxy(phenacyl)- phosphonium Bromide

Solutions of trineopentyl phosphite (3,0 g, 1.0 mol.

equiv.) in acetone (2 ml) and of-bromoacetophenone (2.5 g,

1.22 mol. equiv.) in acetone (6 ml) were mixed at room

temperature. After 1.5 h the acetone was removed under

reduced pressure and anhydrous ether was added. The white

crystalline product that was precipitated was filtered off,

washed with ether, and dried under high vacuum to give

trineopentyloxy(phenacyl)phosphonium bromide (0.6 g, 129^)

(Found: C, 56.8, H, 8.1. C2^H^QBr0^P requires: C, 56.2,

H, 8.29^), m.p. 85 - 86 (sealed tube), (CDCl^) 1.02

(Me^C, s), 4.35 (CH2OP, d, 4.8 Hz), 5.74 (CH^P, d. JPQj^ 18.3 Hz), 7.3 - 8.4 (Ar, m), cTp (CDCl^) 4I.O ppm.

In a similar preparation, trineopentyl phosphite (3.3 g,

1.0 mol. equiv.) and a-bromoacetophenone (2.5 g, 1.11 mol.

equiv. ) were allowed to interact for 3 h. After removal of

acetone and the addition of ether, the mixture was first,

cooled to 0 °C and then maintained at I6 °C overnight to yield crystals of the same intermediate (1.1 g, 20%)

(Found: Br, 16,2. Calc, for C25H^QBr0^P : Br, 16.39^), m.p.

83 - 84 °C (sealed tube).

155 Preparation of Dineopentyloxy(phenacyl)- phenylphosphonium Bromide

(a) Preparation in ether

a-Bromoacetophenone (2.6 g, 1.0 mol. equiv.) in dry

ether (approx. 25 nil) was added from a dropping funnel, with stirring and cooling (16 °C) to dineopentyl phenyl- phosphonite (3.75 g, 1*0 mol. equiv.) which was diluted with 1.0 ml of dry ether. After 10 min the product started

to precipitate, and after 1.5 h crystals were filtered off, washed with ether and dried under high vacuum; yield 0.7 g, m.p. 122 °C (sealed tube). After standing overnight (app­

rox. 18 h) the ethereal residue yielded a second crop of

crystals (1.3 g), m.p. 122 °C (sealed tube). The total

yield of dineopentyloxy(phenacyl)phenylphosphonium bromide

was 2.0 g (31.59^) (Found; C, 59.5, H, 7.1, Br, 16.4,

P, 5.9. C^^H^^BrO^P requires: C, 59.9, H, 7.1, Br, 16.6,

P, 6.496), cTjj (CDCl^) 1.07 (Me^C, s), 4.34 [CH^OP, d AB,

ci(H^) 4.42, d'(Hg) 4.26, J^g 8.0, 4.0, JpocH(B)

4.1 H z7, 5.99 (CH2?, d, Jp^g 15.6 Hz), 7.2 - 8.4 (Ar, m),

dp (CDCl^) 67.1.

(b) Preparation in chloroform

c(-Bromoacetophenone (2.1 g, 1.0 mol. equiv.) was dissol­

ved in dry chloroform (5 ml), and added slowly with shaking

and cooling (16 °C) to dineopentyl phenylphosphonite (3.0 g,

1.0 mol. equiv.), which was diluted in 1.0 ml of dry chlo­

roform. After 2.5 h, dry ether (40 ml) was added to the

reaction mixture. After a further 0.5 h white crystals which

appeared were filtered off, washed with ether and dried under high vacuum to give the phosphonium bromide (1.5 g.

156 29-4%)> m.p. 118 (sealed tube), with the same n.m.r. spectrum as in the previous experiment.

Preparation of Neopentyloxy(phenacyl)- diphenylphosphonium Bromide

(a) Preparation in acetone

a-Bromoaoetophenone (1.5 g» 1*0 mol. equiv.) was dissol­ ved in the smallest possible quantity of dry acetone (app­ rox. 4-5 nil), and was added dropwise to neopentyl diphenyl- phosphinite (2.0 g, 1.0 mol. equiv.) with shaking and cooling (in a water bath at room temperature) since the reaction is strongly exothermic. Within 15 niin the reaction mixture became cloudy and a precipitate appeared. The crys­ tals were filtered off, washed several times with dry ether and then dried under high vacuum to yield the first crop of product (1.36 g, 39.0%) m.p, I46 - 148 (sealed tube).

After a further 4 h a second crop (0.23 g, 6,8%) was iso­ lated. The total yield of neopentyloxy(phenacyl)diphenyl- phosphonium bromide 1.58 g (45*8%) (Found; C, 6 3 .1, H, 6.0,

Br, 16.0, P, 6 .5 . requires; C, 6 3 .7 , H, 6.0,

Br, 1 7 .0 , P, 6 .6%), cTjj (CDCl^) 0.99 (Me^C, s), 4.15 (CH2OP, d, JpQ(.jj 4.2 Hz), 6.25 (CH2P, d, Jp^jj 12.6 Hz), 7-20 - 8.35 ppm (Ar, m), cTp (CDCl^) 68.3.

(b) Preparation in chloroform

a-Bromoacetophenone (1.42 g, 1.0 mol. equiv.) was dissol­ ved in the minimum of dry chloroform (4.5 ml) and was added slowly to neopentyl diphenylphosphinite (1.95 g> 1.0 mol. equiv.) also in dry chloroform (1 ml), with shaking and cooling in ice—water. After 10 min the reaction mixture

157 became cloudy and crystals started to separate. After 3 h the precipitate was filtered off, washed with dry ether and dried under high vacuum to yield the product (2.47 g), m.p. 135 - 136 °C (sealed tube) (Found: C, 53.1, H, 5.1,

Br, 29»096), d^p (CDCl^) 68.3 ppm. The product was recrystal­ lised by dissolving it in dry chloroform and adding dry ether, followed by filtration and drying under high vacuum to give white crystals (stable when stored under CHCl,), m.p. 143 C (sealed tube), of the phosphoniura bromide con­ taining one bound CHCl^ molecule of crystallisation i. e.

Ph2 (RO)P'^CH2COPhBr;CHCl^, (Found; C, 53.1, H, 5.2.

^26^29®^^^3°2^ requires: C, 53.0, H, 5.0%), (3' (CDCl^)

0.99 (Me^C, s, 9H), 4 .15 (CH^OP, d, JpQ^g 4.2 Hz, 2H), 6.25

(CH^P, d, 12.6 Hz, 2H), 7.2 - 8.3 (Ar, m, I6H) inclu­ ding CHCl^ peak at cf 7.25 (s, 1H).

(c) Preparation in ether

o(-Bromoacetophenone (2.3 g, 1.0 mol. equiv.), dissolved in dry ether (60 ml), was slowly added from a dropping funnel to neopentyl diphenylphosphinite (3.0 g, 1.0 mol. equiv.) with stirring at room temperature. An oily liquid began to appear but did not produce crystals after 3 h at room temperature or after 24 H at 8 °C. Then the ether was evaporated off, the oily mixture remaining was dissolved in dry chloroform and the product was precipitated with ether, washed with ether, and dried under high vacuum to give the product as a white crystalline powder containing one molecule of CHCl^ of crystallisation, m.p. I40 - 144

(sealed tube) (Found: C, 54.3, H, 6.0%). The ^H n.m.r.

158 spectrum (CDCl^) showed the same signals as the products prepared above.

The Perkow intermediate

Preparation of Neopentyloxydiphenyl-l-phenvl- vinvlQxvPhosphonium Chloride

Qf-Chloroacetophenone (1.13 g, 1.0 mol. equiv.), dissolved in the minimum of dry chloroform (approx. 2.5 ml), was cooled to between -5 and 0 and then slowly added with shaking and cooling to neopentyl diphenylphosphinite

(1.99 g, 1.0 mol. equiv.). The reaction mixture was concen­ trated under H2O pump pressure and then high vacuum and dry ether (10-15 iiil) was added to give a cloudy, oily liquid, which was washed by decantation with a small por­ tion of ether at low temperature in a nitrogen atmosphere.

The oily mass solidified when stored under ether at -6 °C for two days, after which it was washed with ether at low temperature and dried under nitrogen to give white crys­ tals of neopentyloxydiphenyl-1-phenylvinyloxyphosphonium chloride (Pound: C, 62.95» H, 6.1. C2^H20C1O2P with 0.5 mol CHCl^ requires: C, 62.9, H, 6.1%), (CDCl^) 0.98

(Me^C, s), 4.04 (CH2OP, d, 4.8 Hz), 5.48 (CH2=C,

5.41 - 5 .54, m), 7.20 - 8.20 (Ar, m).Products of decomposi­ tion, i.e. neopentyl chloride cT3.1 (CH2 , s), and vinyl di- phenylphosphinate cT5.15 (CH2=C, m, 5.05 - 5.25 ppm) and traces of unreacted phenacyl chloride cT4.72 ppm were also present, cTp (CDCl^) 55.9. The crystals kept under ether at room temperature were decomposed after one week.

159 III.2.3 OBSERVATION OF THE COURSES OF THE DECOMPOSITION OF ISOLATED PHOSPHONIUM INTERMEDIATES IN DRY CDCl^ BY ^H n.ra.r. SPECTROSCOPY

Decomposition of Arbuzov intermediates

Decomposition of Trineopentyloxy(phenacyl)- phosphoniiim Bromide in Deuterochloroform

The phosphonium bromide (0.1 g) was dissolved in deutero- chloroform (£a. 2*5%) in an unsealed H n.ra.r. tube. After

10 min signals were observed for trineopentyloxy(phenacyl)- phosphonium bromide (87 mole %), cT 1.02 (Me^C, s), 4.55

(CH^OP, d, 4.8 Hz), 5.74 (CH^P, d, 18.3 Hz),

7.5 _ 8.4 ppm (Ar, m), and the decomposition products dineopentyl phenacylphosphonate (13 mole %), cT 0.86

(Me^C, s), 3.72 (CH2OP, d, 4.8 Hz), 3.65 (CH^P, d,

22.8 Hz), 7.2 - 8.1 ppm (Ar, m) and neopentyl bromide

(13 mole %), S 1.03 (Me^C, s), 3.27 Ppm (CH2 , s). Decompo­ sition increased to 30.6% after 25 min at 33 ^C, to 86.0% after a further 15 h at room temperature, and was complete after 3 days.

Thermal Decomposition of Dineopentyloxy(phenacyl)- phenylphosphonium Bromide in Deuterochloroform

The phosphonium bromide was dissolved in deuterochloro- form (oa, 3.0%) and the solution was sealed in a H n.m.r. tube. S ig n a ls for the ketophosphonium intermediate only

160 were observed at S I .07 (Me^C, s), 4.34 (CH^OP, m), 5.99

(CH2P, d, *^PCH Hz), 7*2 - 8.4 ppm (Ar, m). No change r occured in 3 days at room teraperature. The tube was then heated at 100 °C (oil bath) for O .5 h after which the phosphonium intermediate had decomposed completely to two products; neopentyl phenacyl (phenyl )phosphinate (S' 0.81

(Me^C, s), 3.3 - 4.0 (CH2OP and CH2 P, overlapping signals, m), 7.26 - 8.3 (Ar, m), and neopentyl bromide cT 1.02

(Me^C, s), 3.22 (CH«C, s). The tube was then opened and a few drops of D2O were added. After shaking and standing

24 h the H n.m.r. spectrum showed that the signal at

3.77 ppm (PCH2 » dd) had disappeared because of exchange with D. The signal at 3.52 ppm (POCH2 , m) which had pre­ viously overlapped with the PCH2 signals, however remained.

Thermal Decomposition of Neopentyloxy(phenacyl)- dlphenylphosphonium ~Bromi^ in Deuterochloroform

The phosphonium bromide was dissolved in deuterochloro- 1 1 form (ca, 39é) and sealed in a H n.m.r. tube. The H n.m.r. spectrum showed signals for the phosphonium intermediate only at cTO.99 (Me^C, s), 4.15 (CH2OP, d, JpQQjj 4.2 Hz),

6.25 (CH2 P, d, JpQjj 12.6 Hz), 1.2 - 8.3 Ppm (Ar, m). No change occurred in 5 days at room temperature. The tube was heated for 0.5 h at 100 °C in an oil bath, after which time the phosphonium bromide had decomposed completely to phena­ cyl (diphenyl )phosphine oxide^^ cT^4.14 (CH2P» d, JpQjj 15.6 Hz) and neopentyl bromide, (5* 1.02 (Me^C, s), 3.25 (CH2 C, s),

7 .3 - 8.5 ppm (Ar, m).

161 In another experiment solution of the phosphonium bro­ mide was heated in a sealed n.m.r. tube at 80 °C (40 min) (decomposition at this temperature was very slow) and then at 90 °C for 20 min. The decomposition was below

509é under these conditions.

Decomposition of the Perkow intermediate

Decomposition of Neopentyloxy(diphenyl)-1-phenyl- vinyloxyphosphonium Chloride in Deuterochloroform

Isolated crystals of the Perkow intermediate were placed in a n.m.r. tube and dissolved in deuterochloroform

(ca. 4%). The n.m.r, spectrum immediately after prepara­ tion showed signals for the vinyloxyphosphonium interme­ diate (45.7 mole %) S 4.04 (Me^CCH^, d), 5.48 ppm (CH2=C, m), the vinyl phosphinate (54.3 mole %) cT 5.10 (CH2=C, m), and neopentyl chloride S 3»3 (CH^, s), and some minor impurities, mainly phenacyl chloride, S 4.72 ppm.

The tube was kept at 33 °C and as the decomposition pro­ ceeded the signals of the vinyloxyphosphonium compound,

(S' 4.04 (d), 5.48 (m) decreased and the signals for the decomposition products,

The decomposition is summarized in the table below.

Time after preparation phosphonium phosphinate (min) i%) {%)

0 45.7 54.3 15 34.8 65.2 35 24.7 75.3 60 16.6 83.4 The half-life of the decomposition at 33 C was ca. 40 min.

162 IXJ,2,4 preparation a n d i s o l a t i o n o p ARBUZOV AND PERKOW PRODUCTS

Preparation and Isolation of the following Arbuzov Products formed on the Decomposition of the Corresponding isolated Arbuzov Intermediates

Preparation and Isolation of Neopentyl Phenacyl- (phenvl)phosphinate

Dineopentyloxy(phenacyl)phenylphosphonium bromide (1.2 g,

2.5 mmol) was dissolved in deuterochloroform (4 ml) and -1 sealed in three H n.m.r. tubes. The tubes were heated for

40 min at 100-105 °C in an oil bath, after which the

n.m.r. spectrum confirmed that the phosphonium salt had decomposed completely. The contents of the tubes were then combined and concentrated under high vacuum to give a hard white solid (0.7 g) (85.4% calculated as the ketophosphi- nate), m.p. 71-77 °C (sealed tube). The product was re­ crystallised from ether/petroleum (b.p. 40-60 ^C), washed well with petroleum, and dried under high vacuum to give neopentyl phenacyl(phenyl)phosphinate (0.48 g, 58.5%)> m.p. 82-85 °C (sealed tube) (Found: C, 68.8, H, 7*1 •

^19^25^3^ requires: C, 69.1, H, 7*0%), (CDCl^) 0.81

(Me^C, s, 9H), 3.52 [CH2OP, d AB, <5'(H^) 3 .67, <5'(Hg) 3.37,

Ja b JpoCH(A) JpoCH(B) «"h overlapping with 3.77 [CH2P, dd AB, (jr(H^) 3 .79, cT(Hg) 3.74, 0, JpocH(A)

18.4, JpocH(B) 17.5 Hz] (4H total), 7.3-8.3 (Ar, m, 10H),

163 íp (CDCl^) 32.1, m/z 330 (lO^é, M"^), 261 (lOO^é), IR (KBr -1 disc) 1667, 1 2 1 9 , VpQ^ 1031 cm , The filtrate gave a second crop of crystals (0.15 g, 18.396), m.p.

74-79 °C (sealed tube). Total yield was 7 6 .896.

Preparation of Phenacyl(diphenyl)phosphine Oxide

Neopentyloxy(phenacyl)diphenylphosphonium bromide

(0.66 g, 1.4 mmol) was dissolved in deuterochloroform 1 (4.5 ml) and the solution was sealed in three H n.m.r. tubes. The tubes were heated at 100-105 C in an oil bath 1 for 40 min, after which the H n.m.r. spectrum showed that

the phosphonium bromide had decomposed completely. The com­ bined solutions were concentrated under high vacuum to give

a hard white crystalline product (0.4 g, 89.09é calculated

as the phosphine oxide), m.p. 126-134 °G. Recrystallisation

from dry chloroform/ether gave long thin crystals which were washed with ether and dried under high vacuum to give phenacyl(diphenyl)phosphine oxide, m.p. 138-140 °C (sealed

tube) (lit,^^ m.p. 14O-I4O .5 °) (Found: C, 75»5, H, 5.5.

Calc, for 020^17^2^* 75-0, H, 5.496), cTj^ (CDCl^) 4.14 (CHpP, d, JpQjj 15.6 Hz 2H), 7 .3-8.5 (Ar, m, 15H), (Tp

(CDClj) 28.2, m/z 320 (2796, M"^), 201 (100%), IR (KBr disc)

Vc_o Vp^Q 1184 (lit.^"^ 119 0 ) cm*''.

164

^ 4 Preparation and Isolation of the following Perkow Products

Preparation of Neopentyl 1-Phenylvinyl Phenylphosphonate

Phenacyl chloride (5»56 g, 1 mol. equiv.) in dry chloro­ form (15 ml) was added slowly, stirring well, to dineopen­ tyl phenylphosphonite (10.5 g, 1 mol. equiv.) also in dry chloroform (8.0 ml) at 0 to 5 °C. The temperature was then gradually increased to 25 °C and the mixture was stirred for another hour before being stoppered and kept in dry conditions for 2 days. The mixture was then concentrated under reduced pressure giving 1 1 . 5 g (94%) of crude pro­ duct. Distillation of the product gave a first fraction

(4.0 g), b.p. 123-144 °C at 0.1 mmHg, consisting of a mix­

ture of vinyl phosphinate and phenacyl chloride, a second

fraction (5.0 g), b.p. 170-184 °C at 0.1 mmHg, which was mainly vinyl phosphinate with traces of phenacyl chloride,

and a third fraction consisting of pure neopentyl 1-phenyl-

vinyl phenylphosphonate (2.0 g, 17%) b.p. 184-190 at 0.1

mmHg, n^^ 1.5420, (Found; C, 69.1, H, 1.0, C^^H^^O^P re­

quires: C, 69.1, H, 1.0%), (CDCl^) 0.89 (Me^C,s, 9H),

3.77 (CHpOP, d, 2H, JpQCH (CH2=C, d, 2H,

2.5 Hz), 7.2-8.1 (Ar, m, 10H), (Tp (CDCl^) 15.0, PC= CH m/z 330 (9%, M"^), 105 (100%), IR v^^^ 1622, Vp^^ 1260, -I V 990-1040(br.)cm" . The total yield in the second and POC third fractions was 1.0 g, (5 8 .0%).

165 Preparation of 1-Phenylvinyl Piphenylphosphinate

Oi-Chloroacetophenone (2.25 g» 1 mol. equiv.) was dissol­ ved in dry chloroform (4 ml) and was added from a dropping funnel to neopentyl diphenylphosphinite (4.0 g, 1 mol. equiv.) with stirring and cooling at -25 °C; the reaction is exothermic. Dry ether (10 ml) was added to the viscous mass and the mixture was kept at room temperature for 24 hours. Solvent was removed under reduced pressure to leave a viscáis residue which solidified on standing (12 h). The residue was then distilled to give a first fraction

(0.2 g), b.p. 60-120 °C at 0.1 mmHg, which solidified in

the receiver, a second fraction (1.2 g) b.p. 160-190 °C at 0.1 mmHg which also solidified on standing, and the

third main fraction consisting of 1-phenylvinyl diphenyl- phosphinate^^ (1.4 g, 29*8%) b.p. 190-194 °C at 0.1 mmHg

as a very viscous liquid which solidified after 2 hours, m.p. 77-78 (sealed tube) (Found: C, 75*3, H, 5-4.

Calc, for C2q H^^02P: G, 75.0, H, 5-490), (CDCl^) 5.15

(CH2=C, m, 2H), 7.2-8.2 (Ar, m, 15H), (Tp (CDCl^) 29.7, m/z 320 (27%, M"^), 201 (100%), IR (KBr disc) v^^^ 1622, -1 1029, 1005, 990 cm Vp^O ^ POC

166 III.3 PREPARATION AND THERMAL DECOMPOSITION OF THE INTERMEDIATES FORMED BY REACTION OF NEOPENTYL N ,N ,N ',N »-TETRAMETHYLPHOSPHORODIAMIDITE WITH HALOGENOMETHANES

Preparation of Neopentyl Phosphorodichloridite^^*

Neopentyl alcohol (30.0 g, 1.0 mol. equiv. ) in dry ether

(60 ml) was added dropwise (1.5 h) with stirring to phos­ phorus trichloride (50,0 g, I.07 mol. equiv.) in dry ether

(45 ml)* cooling to -1 5 °C (in an ice salt bath). The clear liquid mixture was brought to room temperature, stored for

15 h in dry conditions, and concentrated under reduced pressure. Distillation gave two fractions: 4 1 .0 g, (6 3.79^)» b.p. 64 - 66 °C at 20 mmHg, n^^ I.462O (lit."^^ b.p. 47-5 -

48 °C at 10 mmHg), (ST^ (CDCl^) 0.95 (Me^C, s, 9H), 3.88

(CH^OP, d, JpocH ^P 17 8 .3 , and 5.0 g, (7 .896), b.p. 66 - 67 °C at 20 mmHg, 1.4590; total yield

46.0 g (7 1 .5%).

Preparation of Neopentyl N,N,N*,N'-Tetramethyl- phosphorodiamidite^^

Neopentyl phosphorodichloridite (23.2 g, 1.0 mol. equiv.) in sodium dried petroleum ether (80 ml, b.p. 30-40 °C) was added in small portions (1 h) to dimethylamine (23.0 g,

4.0 mol. equiv.) also in dry petroleum ether at -20 to

-10 °C, stirring well. Vigorous reaction occurred producing white fumes. The mixture was then slowly brought to room

167 temperature, stored over night in dry conditions, filtered under nitrogen, and the filtrate was concentrated under reduced pressure to give a liquid residue (21.6 g, 8 5 .4%).

Distillation gave the pure product (17*8 g» 10,3%), b.p.

86-87 °C at 15 mmHg, 1.4430, (Tg (CDCl,) 0.90 (Me,C,

s, 9H), 2.5 (M62N, d, 8.6 Hz, 6H), 3.19 (CH2O, d.

JpOCH ^-5 Hz, 2H), (Tp (CDClj) 137-2.

Preparation of Neopentyl Phosphorodibromidite 47

Neopentyl alcohol (30.0 g, 1.0 mol. equiv.) in dry

ether (60 ml) was added dropwise (1.5 h) to phosphorus tri­ bromide (94.7 g, 1*0 mol. equiv.) in dry ether (30 ml),

cooled to -I5 to -20 and stirred well. The the reaction mixture was brought to room temperature, stored over night

in dry conditions, concentrated under reduced pressure and

distilled twice to give the pure product (46.4 g, 49.1%),

b.p. 45-50 °C at 0.1 mmHg, (CDCl^) O .96 (Me^C, s, 9H),

3.83 (CH2O, d, JpocH ^P 200.4 ppm.

Preparation of Neopentyl N .N-Dime thy1pho s phor- amidochloridite

Dimethylamine (6.3 g, 2.0 mol. equiv.) in petroleum

ether (b.p. 30-40 °C) (23 ml) was cooled to -20 °C and

added dropwise with stirring (15 min) to neopentyl phos-

phorodichloridite (13.2 g, 1.0 mol. equiv,) also in light

petroleum (47 ml) at -20 °C. The mixture was filtered at

room temperature and the filtrate was distilled after

removal of solvent to give neopentyl N.N-dimethylphos-

phoramidochlorldite {1*2 g, 52.3%), b.p, 44-48 °C at

168 0.1 mmHg, 1.4581 (Found: Cl, I7 .8 . C^H^^CINOP requires:

Cl, 1 7 .9%), cTjj (CDCl^) 0.94 (Me^C, s, 9H), 2.66 (Me2N, d.

12.6 Hz, 6H), 3.55 (CH2O, d, JpQQjj 7.2 Hz, 2H), TNCH ¿Tp (CDCl^) 1 7 8 .4 .

Preparation of Neopentyl N.N-Dimethylphosphor- amidobromidite

Dimethylamine (7 .I g, 2.0 mol. equiv.) in petroleum

ether (b.p. 30-40 °C) (30 ml) was cooled to -20 °C and

added dropwise (20 min) to neopentyl phosphorodibromidite

(21.0 g, 1.0 mol. equiv.), also in light petroleum (60 ml)

at -20 °C. Diethyl ether (50 ml) was added at room tempe­

rature, the mixture was filtered, and the filtrate was

concentrated and distilled to give a main fraction (10.7 g)>

b.p. 55-63 °C at 0.1 mmHg. Redistillation afforded pure

neopentyl N.N-dimethylphosphoramidobromidite (2.8 g, 15.3%),

b.p. 81-82 °C at 0.3 mraHg (Pound: Br, 31.6. C^H^,^BrN0P re­

quires: BPi 33.0%), cTjj (CDCl^) 0.96 (Me^C, s, 9H), 2.64

(Me2N, d, JpjjQjj '•3.8 Hz, 6H), 3.65 (CH2O, d, JpQQjj 7*2 Hz,

2H), dp (CDOl^) 198.3.

Preparation of Bisdimethylamino(methyl)neopentyl- oxyphosphonium Chloride

Neopentyl n ,N,N^N'-tetramethylphosphorodiamidite (4 .O g,

1.0 mol. equiv.) and chloromethane (2.4 g, 2.48 mol. equiv.)

were mixed at -80 °C, sealed in a glass tube, and allowed

to stand at room temperature. After 24 h a little solid

separated and a second liquid layer formed which became

solid in 6 days. Excess of chloromethane was then removed

169 to leave a white solid (5.0 g) which gave a cloudy solu­ tion in CDCl^. Filtration under ai’i’oi’ded tétraméthyl­ ammonium chloride (0,22 g, 10,49é), sublm, > 360 °C, which was identified by infrared (KBr disc). Addition of dry ether to the filtrate gave an oily layer which remained as a liquid residue (4*3 g) after drying under high vacuum and which contained the phosphonium chloride, cfjj (CDCl^)

1.00 (Me^C, a), 2.42 (MeP, d, 14.4 Hz), 2.90 (Me^N, d, 10.8 Hz), 3.90 (CH2O, d, JpQQjj 4.8 Hz), cTp (CDCl^)

59.8 and the phosphoramidochloridite, (C 0.95 (Me,C, s).

2.66 (M62N, d, Jpj,(,jj 9.6 Hz), 3.55 (CH2O, d) (mol ratio ca.

83 ; 1 7 ), cTp 1 7 9 .2 . The product crystallised in 10 weeks

at 0 °C and was then washed with anhydrous ether (0 °C)

and dried under high vacuum to give bisdimethylamino-

(methyl)neopentyloxyphosphonium chloride (1.5 g, 30.19^),

m.p. 55-60 °C (sealed tube) (Found: C, 45.8, H, 10.2, Cl,

13.5, N, 11.3, P, 11.3. C,,q H25C1N20P requires; C, 46.8, H,

10.2, Cl, 13.8, N, 10.9, P, 12.1°/), (CDCl^) 1.0 (Me^C,

s, 9H), 2.37 (MeP, d, Jp^^ 14.7 Hz, 3H), 2.89 (Me2N, d.

Jpj^CH 1^*2 Hz, 6H), 3.86 (CH2O, d, JpQQjj 5.0 Hz, 2H), cf^

(CDCl^) 60.3 .

Preparation of Bisdimethylamino(methyl)neopentyl- oxyphosphonium Bromide

Neopentyl N,N,N',N’-tetramethylphosphorodiamidite

(1.7 g, 1.0 mol. equiv.) was added to bromomethane (2.0 g,

2.5 mol. equiv.) at -80 °C and the mixture was allowed to

stand at room temperature (22 h). After removal of the

excess of bromomethane the solid residue (2.8 g) was

170 dissolved in the minimum of CDGl,. (some cloudiness remai- ned) and shown to contain the phosphonium bromide, 1.01

(Me^C, s), 2.44 (MeP, d, 15.3 Hz), 2.93 (Me2N, d,

J 10.2 Hz), 5.93 (CH2O, d, JpQçjjj 4.8 Hz), cTp 60.2, and the phosphoramidohalidite, 6'j^ 3.65 (CH2O, d, JpQQjj 7*2 Hz)

(mol ratio ca. 86 ; I4 ), 199.1. Solid which separated was washed with chloroform and ether and identified

(infrared, KBr disc) as tétraméthylammonium bromide

(0.05 g, 3 .9%), m . p . > 300 °C. Concentration of the fil­ trate followed by addition of dry ether gave white, crys­ talline bisdimethylamino(methyl)neopentyloxyphosphonium bromide (2.2 g, 88.6%), (Found: C, 38.3, H, 8 .4 , Br, 26.5,

N, 9 .5 , P, 10.0. C^QH25BrN20P requires: C, 39.9, H, 8.6,

Br, 26.6, N, 9.3, P, 10.3%), m.p. 122 °C, (CDCl^) 1.01

(Me^C, s, 9H), 2.47 (CH^P, d, Jp^H ^5.6 Hz, 3H), 2.92

(Me2N, d, ‘^pjiQjj 6H), 3.93 (CH O, d, JpQQpj 4.8 Hz, ^PNCH 2 2H), cTp (CDCl^) 60.2

Preparation of Bisdimethylamino (methyl )neopentyl- oxypho3phonium Iodide ®

Neopentyl ,Ut-tetramethylphosphorodiamidite 31 (1,0 g, 1,0 mol. equiv.) in CDCl^ was placed in a ^ P n.m.r

tube and cooled in an ice-bath. lodomethane (1.5 g, 2.18

mol. equiv.) was added slowly, when a vigorous exothermic

reaction occurred. Immediately after the addition the

^^P n.m.r. spectrum showed a signal due to the phosphonium

iodide, cTp 59.7 ppm. No other products were detectable.

Then the deuterochloroform solution was treated with dry

ether, which precipitated white crystals of the phosphonium salt.

171 Investigation of the Action of Halogenomethanes on N.N-Dimethylphosphoramidohalidites

(a) Chioromethane

The phosphoramidochloridite (2.5 g» 1*0 mol. equiv.)

and chloromethane (1.7 g» 2.66 mol. equiv.) were mixed at

-80 °C and sealed in a glass tube. After 28 h at room

temperature the tube was cooled and opened and chloro­ methane was allowed to escape. The residue (2.8 g) contai­

ned only the unreacted chloridite, (CDCl^) 0.95 (Me^C,

s), 2.66 (Me2N, d, JpjjQjj 12.6 Hz), 3.56 (CH2O, d, JpQQjj

7.2 Hz) and a little chloromethane (Tjj 2.99(s). Signals

assignable to the dichloridite, R0PCl2f 6^p (GDCl^) 178*4-

were absent.

(b) Bromomethane

By a similar procedure the phosphoramidobromidite

(1.7 g, 1.0 mol. equiv.) and bromomethane (2.0 g, 3.0 mol.

equiv.), after 4 days at room temperature, gave a residue

(2.1 g) containing the unreacted bromidite, c5^ (CDCl^)

0.97 (Me^C, s), 2.63 (Me2N, d, 13.5 Hz), 3*64 (CH2O, d, J 7.2 Hz) and some bromome thane, 6*^^ 2.63(s). No

dibromidite, R0PBr2> <^p (CDCl^) 198.1 was detectable.

Thermal Decomposition of Bisdimethylamino(methyl)- neopentvloxyphosphonium Halides

(a) In anliydrous CDCl^ Solutions of the halides (ca. 10^) in CDCl^ were heated

in sealed '’r n.m.r. tubes at 100 °C. In each case, decompo-

sition gave the neopentyl halides (ci^ as reported)^^ and

172 methylphosphonic bisdimethylamide, djj 1.4 (MeP, d, JpQj^

13.2 Hz, 3H), 2.60 (M02N, d, 10.8 Hz, 12H), d'p 36.9

(lit.dp 58.0) as follows Cl, 60/4, 94/9.5,

99/21; Br, 80/3.5, 98/7.5, 100/13.5.

(b) In the presence of water

A solution of the phosphonium bromide (0.1 g) in CDCl^

(1 ml) was shaken with water (0.2 ml) and the mixture was then heated for 44 b at 100 °C (reaction complete) in a •1 sealed tube. H n.m.ir. showed the presence of neopentyl bromide, (Tl.03 (Me^C, s), 3.27 (CH2 , s), methanephospho- nic bisdimethylamide, cT 1 .40 (MeP, d, JpQ^ 15.0 Hz), 2.61

(Me2N, d, JpjjQjj 10.0 Hz), and the dimethyl ammonium ion, cT 2.57 (s) (ca. 18 mole %), In two further experiments the bromide (0.58 g) was heated (44 h at 100 *^C) with water

alone: (a) 0.06 g ^mol. equiv.); (b) 0.12 g (4 mol. equiv.)

Tube (a) formed three layers, the upper layer slowly depo­

siting white needle-like crystals of dimethylammonium bro- + mide (Tjj (CDCl^) 2.69 (Me, s, 6H), 9.56 (NH2 , br s, 4H).

The middle layer was dissolved in CDCl^ and treated

successively with ether and acetone to give an oily liquid

which crystallised in 8 weeks. The crystals were washed

Hith ether, dried, and identified as dimethylammonium

neopentyl methanephosphonate, m.p. 95-100 °C, cfjj (CDCl^)

0.91 (Me,0, a, 9H), 1.24 (MeP, d, Jp^jj 15.9 Hz, 3H), 2.53

(MeigNHg, a, 6H), 3.45 (OHpO, d, Jp^^jj 5.1 Hz, 2H),

(CDCl,) 24.0 (quartet of triplets), ¿'^ (CDClj) 11.8 (MeP,

d, Jp(j 137.5 Hz), 26.0 (MejC, a), 32.0 (Me^C, d, Jp^pg + 7.3 Hz), 34.5 (MepHHp, a), 74.5 (CHpOP, d, Jp^^ 6.1 Hz).

173 Tube (b) formed two layers from which volatiles were remo­ ved under reduced pressure. The residue was dissolved in

CDCl,, and treated with ether/acetone to give a first crop of crystals which were washed with ether and dried to give bisdimethylammonium methanephosphonate (0.15 g)> m.p. 60 °C, + djj (CDCl^) 1.30 (MeP, d, 16.2 Hz, 3H), 2.66 (Me2NH2 , s,

12H), 9.89 (NH2 , br s, 4H), (fp (CDCl^) 23.5 (quartet),

A further crop of crystals consisted of dimethylammonium neopentyl methanephosphonate, m.p, 90-100 °C, with n.m.r.

data as given above.

Absorption of Water by the Phosphonium Chloride

The chloride (0,18 g) was left open to the atmosphere

when it absorbed water (0.033 g, 2.7 mol. equiv.) in 5 b,

forming a liquid mixture. On standing in a desiccator with

p^O^ overnight the water content fell to 0.023 g (1.9 mol. A equiv.). The H n.m.r. spectrum (CDCl^) showed that no

reaction had occurred. A signal which slowly appeared at

(T2.66 (Me2NH2 ) showed that ca. 27% hydrolysis occurred in

the CDCl, solution in I6 weeks, 3

174

r i REFERENCES

1. Gerrard, W., and H.R. Hudson, Organic Derivatives

of Phosphorous Acid and Thiophosphorous Acid, in

’’Organic Phosphorus Compounds”, Vol. Chapter 13,

p. 7 0 , G.M. Kosolapoff and L. Maier, ed. John Wiley,

New York, 1973»

2. Michaelis, A., and R. Kahne, Chem. Ber., 3 1 , IO48

(1898).

3 . Arbuzov, A.E., J. Russ. Phys. Chem. Soc., 38, 687

(1906). 4 . Arbuzov, A.E., G. Kamai and L.V. Nesterov, Trudy

Kazan. Khim. Tekhnol. Inst., 1_6, 17 (1951); C.A. ¿1_,

5720f (1957). 5 . Arbuzov, A.E., and P.V. Nesterov, Dokl. Akad. Nauk

SSSR, 92, 57 (1953); C.A. 10538b (1954).

6. Arbuzov, A.E., and L.V. Nesterov, Izv. Akad. Nauk

SSSR. Ser. Khim., 1954, 36l; C.A. 954b (1955).

7 . Landauer, S.R., and H.N. Rydon, Chem. Ind. (London), 1951, 313. 8. Landauer, S.R., and H.N. Rydon, J. Chem. Soc., 1953,

2224. 9 . Abramov, V.S., Khim. Primenenie Fosfororgan.

Soedinenii. Akad. Nauk SSSR, Trudy 1-oi Konferents.,

1 9 5 1, 71; C.A. ¿2, 240b (1958). 10. Abramov, V.S., and G. Karp, Dokl. Akad. Nauk SSSR,

2ii 1095 (1953); C.A. 18, 9906g (1954).

175 11. Abramov, V.S., and G. Karp, Zh. Obshch. Khlm., 24.

1823 (1954); C.A. 13887c (1955).

12. Abramov, V.S., and A.P. Rekhman, Zh. Obshch. Khim.,

26, 163 (1956); C.A. ¿0, 13723h (1956).

13. Kosolapoff, G.M., "Organophosphorus Compounds”,

Chapter 11, pp. 325-326, John Wiley, New York, 1950.

14. Kosolapoff, G.M., "Organophosphorus Compounds”,

Chapter 11, p. 330, John Wiley, New York, 1950.

1 5 . Weekes, J.E., Ph.D. Thesis (University of London)

pp. 35-36, 74, 134 (19 72). 1 6 . Kosolapoff, G.M., "Organophosphorus Compounds”,

Chapter 7, p. 121, John Wiley, New York, 1950.

1 7 . Kosolapoff, G.M., "Organophosphorus Compounds",

Chapter 6, p. 102, John Wiley, New York, 1950.

18. Koaolapoff, G.M., »• Organophosphorus Compounds",

Chapter 7 , p. 122, John Wiley, New York, 1950.

1 9 . Arbuzov, A.E., J. Russ. Phys. Chem. Soc., 42 .

395 (1910).

20. Prank, A.V., Phosphonous Acids and Derivatives in

"Organic Phosphorus Compounds", Vol. 4, Chapter 10,

pp. 3OI-303, G.M. Kosolapoff and L. Maier, ed. John

Wiley, New York, 1972.

21. Frank, A.V,, Phosphonous Acids and Derivatives in

"Organic Phosphorus Compounds", Vol. 4, Chapter 10,

p. 301, G.M. Kosolapoff and L. Maier, ed. John

Wiley, New York, 1972.

22. Kamai, G., Dokl. Akad. Nauk SSSR, 66, 389 (1949);

C.A. ii, 127c (1950).

176 23. Kamai, G., and E.H. Gerasimova, Trudy Kazan. Khim.

Tekhnol. Inst.. 1^, 26 (1950); C.A. ¿1, 112731

(1957). 24. Nesterov, L.V., and R.I. Mutalapova, 2h. Obshch.

Khlm.. 21, 1847 (1967); C.A. 68, 38895n (1968).

25. Arbuzov, A.E., G. Kamai, and L.V. Nesterov, Trudy

Kazan, Khim. Tekhnol. Inst., 16. 17 (1951); C.A. ¿1_,

5720f (1957).

26. Arbuzov, A.E., Dissertation, 1905; A., et al., Trudy

Kazan. Khim. Tekhnol. Inst., 16. 17 (1951); C.A. 49 .

9541h, (1955 ).

2 7 . Landauer, S.R., and H.N. Rydon, J. Chem. Soc.. 1954,

2281.

28. Rydon, H.N., and B.L. Tonge, J. Chem. Soc., 1956,

30-4 3. 2 9 . Frank, A.V., Phosphonous Acids and Derivatives in

"Organic Phosphorus Compounds", Vol. 4, Chapter 10,

pp. 309-311, G.M. Kosolapoff and L. Maier, ed. John

Wiley, New York, 1972.

30. Kosolapoff, G.M., "Organophosphorus Compounds",

Chapter 3, p. 50, John Wiley, New York, 1950.

3 1 . Hudson, H.R., A.R. Qureshi, and (Mrs.) D. Ragoonanan,

J. Chem. Soc.. Perkin I. 1972, 1595.

3 2 . Frank, A.V., Phosphonous Acids and Derivatives in

"Organic Phosphorus Compounds", Vol. 4, Chapter 10,

p. 309, G.M. Kosolapoff and L. Maier, ed. John Wiley,

New York, 1972. 3 3 . Gerrard, W., and H.R. Hudson, Organic Derivatives of

177 Phosphorous Acid and Thiophosphorous Acid, in

"Organic Phosphorus Compounds", Vol. 5, Chapter 15,

p. 53, G.M. Kosolapoff and L. Maier, ed. John Wiley,

New York, 1973.

34. Gould, E.S., "Mechanism and Structure in Organic

Chemistry", p. 428, John Wiley, New York, 1959.

35. Hudson, H.R., Quasi-Phosphonium Intermediates and

Compounds in "Topics in Phosphorus Chemistry", Vol.11,

Chapter 5, p. 397, ed. John Wiley, New York, 1983.

36. Sigal, I.S., and F.H. Westheimer, J. Am. Chem. Soc.,

101. 5329 (1979).

37. Chopard, P.A., V.M. Clark, R.P. Hudson, and A.J.

Kirby, Tetrahedron, 2 1 . 1961 (1965).

58. Koziara, A., B. Mlotkowska, P. Majewski, B. Sledzinski,

and A. Zwierzak, Pol. J. Chem., 55. 339 (1981 ).

39. Mlotkowska, B., P. Majewski, A. Koziara, A. Zwierzak,

and B. Sledzinski, Pol. J. Chem.. 55. 631 (1981).

40. Marquading, D., P, Ramirez, I. Ugi and P. Gillespie,

Angew, Chem, Int. Edn., 12. 91 (1973).

41. Grim, S,0., W. McFarlane, E.P. Davidoff, and T. J.

Marks, J, Phys. Chem., 70 . 581 (1966).

42. Murray, M . , and R. Schmutzler, J. Chem. Soc., (B),

1971. 1715.

43. Haake, P,, W. B. Miller, and D.A. Tysser, J. Am.

Chem. Soo.. 86, 3577 (1964).

44. Arbuzov, B.A., N.A, Polezhaeva, V.S. Vinogradova, and

A.K, Shamsutdinova, Izvest. Akad. Nauk SSSR, Ser.

Khim.. 1965, 669; C.A. 2998f (1965).

178 45. Borowitz, I.J., K.C. Yee, and R.K. Crouch, J, Qrg.

Ghem ., I713 (1973).

46. Gerrard, W., A. Nechvatal, and B.M. Wilson, J. Chem. Soc., 1950, 2088.

4 7 . Hudson, H.R., J. Chem. Soc. (B), 1968, 664.

48. Qureshi, A.R., Ph.D. Thesis (University of London)

1982.

49. Bottger, G.L., and A.L. Geddes, Spectrochim. Acta,

21, 1701 (1965).

50. Jones, R.A.Y., and A.R. Katritzky, Angew. Chem., 7 4 .

60 (1962).

5 1 . Vogel, A.I.,”A Text-book of Quantitative Inorganic

Analysis»', p. 575, Longman, London, 3rd Ed., I96I.

5 2 . Roberts, J.D., end Marjorie C. Caserio, "Basic

Principles of Organic Chemistry", p, 1302, W.A.

Benjamin, Inc., Menlo Park, California Reading,

Massachusetts, USA, second Ed., 1981.

5 3 . Karsh, H.H., Phosphorus Sulfur, 12, 222 (1982).

5 4 . Crutchfield, M.M., C.H. Dungan, J.H. Letcher,

V. Mark, and J.R. Van Wazer, "Topics in Phosphorus

Chemistry", p. 227, John Wiley, New York, I967.

5 5 . Maciel, G.E., and R.V. James, Inorg. Chem., 2.,

1651 (1954).

56. Nesterov, L.V., R.I. Mutalapova, S.G. Salikhov, and

E.I. Loginova, Izv. Akad. Nauk SSSR, Ser. Khim., 1971, 414. 5 7 . Hudson, H.R., Quasi-Phosphonium Intermediates and

Compounds in "Topics in Phosphorus Chemistry", Vol.11,

179 Chapter 5» P* 394, ed. John Wiley, New York, 1983

refers to: Nesterov, L.V., A.Ya Kessel, Zh. Obshch.

Khi m ., 37 » 728 (1967); Lewis, E.S., and K.S. Colle,

J « Org. Chem., 4 6 . 4369 (1981 ).

58. Gardner, J.N., and J. Kochiing, U.S. Patent Office,

3,847,999 (1974).

59» Crutchfield, M.M., C.H. Dungan, J.H. Letcher,

V.Mark, and J.R. Van Wazer, »'Topics in Phosphorus

Chemistry", p. 259, John Wiley, New York, 1967;

Riess, J.G., and J.R. Van Wazer, Am. Chem. Soc.. 8 8 ,

2166 (1966).

60. Petnehazy, I., G. Szakal and L. Toke, 1st Actes

Cong. Int. Composes Phosphores, 555 (1977); L. Tôke,

I. Petnehazy and G. Szakàl, J. Chem. Res. (S) 155

(1978).

61. Worms, K.H., and M.Schmidt-Dunker, Phosphonic Acids

and Derivatives, in "Organic Phosphorus Compounds",

Vol. 7, Chapter 18, p. 61, G.M. Kosolapoff and

L. Maier, ed. John Wiley, New York, 1976.

62. Michaelis, A., Annalen. 326. 129 (1903).

63. Hudson, H.R., R.G. Rees, and J.E. Weekes, J. Chem.

Soc., Perkin Trans. 1, 1974, 982.

64. Hudson, H.R,, A. Kow, and J.C. Roberts, J. Chem.

Soc.. Pepsin Trans. 2, 1983, 1363.

65. Henriok, K,, H.R. Hudson, E.M. McPartlin, and

J, Nasiryn, unpublished work.

180 FIGURES AND TABLES

Tables pages

I 100-102

II 104-105

III 106

IV 107

V 112

VI 127

VII 137

VIII 138

IX 141

X 145

XI 150

XII 151

XIII 159-161

XIV 67

XV 70

XVI 72

XVII 78

181 \ \ n' ,■

I f Attention is drawn to ^ the fact that the i t copyright o f this thesis rests with its author. This copy o f the thesis has been supplied ‘ on condition that anyone who consults it is understood to recognise that-its copyright rests with its author and that no quotation from the thesis'^ and no information derived from k may be published without the author’s .‘prior written consent. 067043^86 END